#does nobody know luke brandon

[xoxzso]

[miafic]

can you write something about one of the boys saying their safety level is 0?? maybe luke or shane :(

“Okay, gentlemen,” Lucas said as the boys ran down the stairs. “Safety levels?”

“Ten!” came three shouts, but embedded in them was a voice cheerfully saying, “One!”

Lucas froze. “What?” 

The kids piled onto the couches. 

“Morgan, what did you say?” Lucas asked, staring directly at him.

“What?” Morgan muttered, looking guilty. “Nothing?”

“Up, please. Now. Into the office.” Lucas raised his voice but didn’t break eye contact. “Brandon! Zakk!”

“Lucas, I’m fine-” Morgan protested weakly

“Zakk!” Lucas called again. His voice was even, but it was taking all of his focus not to let it tremble. 

Zakk popped into the room with Bran right behind him. 

Lucas tossed the clipboard onto the couch cushion beside himself and stood, just barely glancing away from Morgan. “You remember that there’s laundry in the wash?” he intoned. A code.

“Uh - y-yeah. I’m gonna get it later.” Zakk looked from Lucas to Morgan and back.

“Okay. Keep your ears open.” 

Zakk nodded. 

“Morgan, come on. Office.” 

“I said I’m fine,” Morgan said.

“Now.” 

The other kids were staring, but Morgan sighed dramatically and obeyed. He threw himself onto the couch as Lucas quickly shut the door. 

“What’s going on?”

“Nothing.”

“No.” Lucas focused hard on keeping his cool. “A safety level of one is very, very serious.”

“I was just kidding,” Morgan tried, but Lucas could see right through him. 

He shook his head slowly. “Tell me what’s going on.” 

“Nothing.”

“I can’t help you unless I know. And you said it out loud, which means you want help. Some part of you does. So let me help you.” Lucas sat down in the desk chair. He softened his voice. “Morgan, let me help you. Nobody has to know anything.” 

The gentleness was working. Morgan’s expression was already changing from false defiance and carelessness to one of vulnerability. 

“I’m not gonna tell anybody. Not even Zakk if you don’t want.”

“I don’t,” Morgan whispered.

“You don’t want me to tell Zakk?”

Morgan shook his head.

“Okay. I won’t. What you say is just for me to know.” 

Morgan bit his thumbnail and let his eyes go unfocused on the carpet. 

“Do you want to talk freely, or do you want to answer some questions?”

“Questions,” Morgan whispered. 

Lucas nodded. “Okay. If I start going in a direction you don’t like or want, just tell me.” 

Morgan nodded, too.

“A safety level of one means that you have a plan to hurt yourself or somebody else and that you’re not going to tell me before you carry it out,” he stated, his chest tight. “Is that an accurate reflection of how you feel?” 

Morgan was still for a long moment, but he finally muttered, “Yeah.”

“Okay. Thank you for being honest with me. Are you wanting to hurt yourself?”

Another nod. And then came the tears. 

Lucas passed Morgan a box of tissues but otherwise sat still in his chair, watching for the slightest hint of trouble. 

“I don’t want to be here,” Morgan said with a sob. “I want to go home. I want my parents to love me, and they don’t, and it’s not fair. I didn’t even get a goddamn chance to try to live on my own before the cops picked me up.” 

Lucas just listened. 

“I miss my friends. I miss school. I hate saying grace and going to church. I hate having a roommate.” Angrily, he shoved the tissue box onto the floor. “I want to go home!” 

“Our goal is to get you home,” Lucas told him. “But we can’t send you there until we know that it’s safe for you, and right now, we don’t know that.”

“Cause my dad’s a fucking homophobe,” he spat. 

“What about your mom?” 

“She agrees with whatever my dad says.” Morgan wiped at his eyes with his fingers, but the tears were still falling quickly, so there wasn’t really much of a point. 

Lucas sighed sadly. “I’m sorry, Morgan.” 

“I wanna see Chandler and TJ,” he said through his tears. “Forest and Lexi and Paige and Tom…” He turned sideways on the couch and kicked his legs out in front of him. 

“How are you planning to hurt yourself?” 

“I took a knife from the kitchen,” he muttered. 

“Is it on your person?” Lucas asked.

Morgan shook his head. “No. It’s in my room.” 

“Where?”

“Under my bed.”

“Thank you,” Lucas said again. Toward the door, Lucas evenly called, “Zakk!” Back to Morgan, he asked, “And do you want to do with it?” 

Breathlessly, Zakk flung the door open. He took in the scene in front of him, relaxed a bit, and then said, “What’s up?” 

“Come in. Close the door.” 

Morgan was clenching his teeth and staring hard at the wall, but Zakk saw the tears on his face and gave his typical, sad, “Oh, dude…” and went over to give him a hug. Morgan leaned into him and started to cry all over again. 

“Okay, you stay here. I’ll be right back,” Lucas muttered. “Hey,” he said when Zakk didn’t look over. “He just gave me a one on his safety level, so.” 

Zakk nodded, his eyes worried. “Morgan,” he murmured, hugging Morgan tighter.

“I’m sorry,” Morgan sobbed. 

“No, heyyy, it’s okay… What happened, huh?”

Lucas slipped through the door. 

“Is Morgan okay?” Noel asked worriedly. 

“Focus on the meeting, please,” Lucas said, and he crossed the space and walked up the steps, ignoring the feeling of all of their eyes on him.

“Uh, Jordan, did you go yet?” Bran asked awkwardly.

“No…”

Lucas went into the bedroom to the right of the staircase and crouched down beside Morgan’s bed. Teddy’s bed, Beau’s bed, Luke’s bed… He swallowed thickly. Awsten’s bed, he forced himself to continue. Calum’s bed. He thought of Teddy’s green coat, of Beau’s big boots, of Awsten’s light purple hair, of Calum’s signature smirk. With a shaky breath, Lucas knelt down mere feet away from where he’d found Luke sprawled on the floor. He lifted the edge of the quilt and looked beneath it - and there was the cheap silverware. It was too dull to do much of anything, but with the right intent, it could do some serious damage. Not as swiftly as Luke’s pocketknife, but-

“Lucas?” Bran called. “Zakk has a question. He says it’s not urgent.” 

“Alright, just a second!” Lucas called back. He grabbed the knife, stood up, and lifted his shirt. He tucked the cold metal into the waistband of his underwear, right against his hip, where he kept it hidden until he could cross the landing and put it inside his own bedroom. Afterward, he jogged downstairs and went back into the office. Morgan was sitting up, sniffling but no longer crying, and Zakk was going through the file cabinet. 

“Other side,” Lucas said shortly. “Third drawer down.” He paused while Zakk muttered his thanks and moved the chair to face the other side of the desk. Lucas kept an eye on Morgan while Zakk’s back was turned. “You’re looking for a safety contract, right?”

“Yep.” 

“Yeah. Third drawer.”

“You’re the best,” Zakk told him with a grateful smile. 

Lucas just shrugged. “Morgan, how are you doing now?” 

“A little better,” he mumbled.

“Good. Zakk, are you gonna take it from here?” 

“Yeah, if that’s okay with Morgan.” 

“It’s fine.” 

“Okay. I’ll be upstairs then,” Lucas said. He inhaled quietly and then said softly, “I kind of want to go swimming.” 

Zakk nodded. “Of course,” he said, and Lucas nodded back, acknowledging that his code had been understood. He didn’t want to swim; he wanted a break. He was going to lie down. 

“Yell if you need me,” Lucas instructed, and he left without another word. “Bran, can you get them in bed when you’re done?” he asked, not bothering to look at them as he passed. “Gentlemen, behave for Bran. There will be consequences if you don’t.” 

There was a murmur of “okay”s, and then Lucas was up the stairs and in his bedroom, closing the door and sitting shakily down on his bed without bothering to turn on the light. The knife was still sitting on his dresser, so he got up and moved it into the top drawer so it was out of sight before going back to sit on his bed.

He couldn’t stop picturing Luke’s massive gash, and he let out a soft moan of pain. He curled up on his side and shut his eyes, hoping that later, Zakk would come check on him. And after about forty-five minutes, he did.

“Morgan’s asleep,” Zakk reported once he’d let himself in and sat down on the side of Lucas’ bed. “I’m not gonna take him to the emergency room tonight. I think he’ll be okay. It sounds like he was just crying for help.” He sighed. “Which is good. But I can tell that he freaked you the hell out.” 

Lucas didn’t move, didn’t speak. 

Zakk leaned down and hugged him. “How are you doing?” 

“Not good,” Lucas whispered. He rolled onto his back so he could see Zakk. “He took a knife from the kitchen and hid it under his bed.” 

“I know. He told me.”

Lucas gulped. “I haven’t been on the floor in there since…” 

“You hardly even go in anymore,” Zakk agreed in sympathy. 

Lucas turned back onto his side. 

“Go ahead and get some sleep,” Zakk told him. “I’ll take care of the paperwork.” 

“Are you sure?” Lucas asked halfheartedly, but he already knew he wasn’t getting back up. “There’s a lot more now tonight.”

“Yeah, I’m sure.” 

He nodded tiredly. 

Zakk leaned back down to embrace him again. “Do you wanna talk about it?”

“No.” 

Zakk nodded. “Okay.” But he didn’t move.

Lucas was quiet for ten seconds before he asked, “What are you doing?” 

“Making sure you get to sleep.” His tone softened. “I know you really wanted to go swimming.” 

Normally, Lucas would have smiled, but Zakk was right. And honestly, he was glad that Zakk was staying. He closed his eyes. As Zakk started rubbing his arm, Lucas’ heart rate began to slow. 

For a long time, the room was silent. The house was silent. 

It had been silent when Luke had-

Lucas bolted up, sending Zakk tumbling. “What if Morgan-”

“Bran’s sitting right outside the doorway,” Zakk soothed. He took hold of Lucas’ upper arms and gently pressed him back into the pillows. “Everything is taken care of. Go to sleep.” 

“Zakk,” Lucas pleaded.

“Shh…” Zakk draped himself over Lucas yet again. 

It took almost half an hour, but Lucas finally fell into uneasy dreams. 

[yunsangelic]

captaindboss’ Hottest NHL Players Survey Responses

I’m demonkonecny bc it’s halloween!!! Happy Halloween!!! Anyway I’m finally posting the results of my hottest players per team survey, (it’s closed now so u can’t take it anymore, sorry) which included ur fav ugly hots like jack eichel and connor mcdavid. Y’all had some colorful write-in responses for me, I included my favorites! Anyway, as not to clog dashes I put it under a read more. If you have questions about how I compiled this data or how I organized it, feel free to ask! Also some of y’all didn’t put an answer for like half the teams???? who raised you.

Montreal Canadians

Carey Price (55.02%--126 of 229 votes)

Jonathan Drouin (23.58%--54 of 229 votes)

Shea Weber (13.10%--30 of 229 votes)

Other* (8.30%--19 of 229 votes)

*= Inconclusive results.

Write-ins

“PK Subban...oh wait...Lars Eller... Oh wait...Drouin...oh wait...Alex Galchenyuk...oh wait shit fine Shea Weber”

“Everyone who has escaped”

“their ‘attitude problem’“

Boston Bruins

Brandon Carlo (30.26%--69(lol) of 228 votes)

Patrice Bergeron (Cause y’all would kill me if I didn’t put him) (29.82%--68 of 228 votes)

David Pastrnak (yum i lov carb) (25.88%--59 of 228 votes)

Other* (14.04%--32 of 228 votes)

*= Inconclusive results.

Write-ins

[About Carlo] “He’s  a baby but also like has an ageless vampire quality which appeals to my ovaries, long conditioned by teen vampire novels”

“Brad Marchand's tongue (only the tongue)” [this ain’t it chief]

“I love my alien father tuukka rask” [r u ok]

Bonus, cause I’m weak:

“no one THINKS pasta is hot COME ON i hate us” [it’s okay, he is VERY hot, that’s why I put him lmao]

“Zdeno chara babey” [R U OK]

Bl*ckh*wks

Girl as if (44.80%--99 of 221 votes) 

Jonathan Toews (22.62%--50 of 221 votes)

If you put pk*ne here i’ll come to your house and murder you*^/other (17.65%--39 of 221 votes)

hahahahaHAHAHA (14.93%--33 of 221 votes)

*= tie between Nick Schmaltz and John Hayden.

^= 3 people want me to come to their houses and murder them, unfortunately it’s still illegal to do so, therefore I will not be doing that.

Write-ins

“toews player portrait makes him look like a human condom”

“i live in chicago and am willing to take 1 for the team and take out k*ne” 

“Bitch you funny but also Alex Debrincat”

New York Rangers

Brady Skjei (46.32%--107 of 231 votes)

Henrik Lundqvist duh (31.17%--72 of 231 votes)

Brett Howden is the right answer despite not being on the roster yet^ (11.69% (lol)--27 of 231 votes)

Other* (10.82%--25 of 231 votes)

*= Inconclusive results.

^= funny enough, like 2 days after I made this Brett made the final cut lmao.

Write-ins

“Chris Kreider (have you seen the golf pic???) [LMFAO yeah I have (it’s probably too NSFW if ur in public fyi if u wanna google it)]

“i'm horny for like half the gd rangers roster!!!!” [lol girl I know, y’all actually have a few cuties I was a lil shocked]

“this is a hot team too bad they suck”

Toronto Maple Leafs

Y’all are demons okay Nikita Zaitsev is a fuckin gem idk why I was surprised about this but I was lmao.

Other* (39.37%--87 of 221 votes) [just know that I hate u all :/]

William Nylander (25.79%--57 of 221 votes)

Nazem Kadri (24.89%--55 of 221 votes)

Nikita Zaitsev (9.95%--22 of 221 votes)

*= Freddie Andersen. 

Write-ins

“william nylander isn't a leaf, firstable, and second it's motch murner” [sjdhkdlsjdj everything about this]

“i'm putting rich clune even tho he's on the marlies. SOMETIMES HE COMES UP. he could benchpress ever leaf on the roster.” [ur valid, when u sent this I was like “FUCK they’re right.”]

“None they look like 25 year olds who smoke crack in the parking lot” [this is low-key mean but I still laughed, cause yeah, white dudes. But I’m not condoning drug abuse or jokes about drug abuse, as this person had no intention of doing, I’m sure. Just wanted to put that because I know some people might be concerned.]

Bonus, again, weak:

"Jxhn Txvxrxs” [jhkhfoij why did u censor his name sis??]

“nobody’s attractive on the leafs” [this isn’t true but I’m petty and it’s funny.]

Detroit Red Wings

I was so fucking offended by some of the dylan larkin SLANDER up in these write-ins, y’all can come to my house and fight me thx.

Dylan Larkin (48.23%--109 of 226 votes)

Henrik Zetterberg (im crying) (31.42%--71(CRYING) of 226 votes)

Other* (11.95%--27 of 226 votes)

Andreas Athanasiou (8.41%--19 of 226 votes)

*= 12 votes for “No one/Not Dylan Larkin” (yall r annoying lmfao), 10 votes for Filip Zadina (he’s a CHILD how dare u)

Write-Ins

“Luke glen denting is hot look at his arms and he’s not too old for ME” [girl when I tell u this shit killed me, I mean I SQUAWKED a laugh out and sent it to the fps gc, I was DEAD] 

“I don't know what any of the red wings look like and it's probably better that way” [????????????????]

“ion know anyone on the wings except zadina and he scored a gwg against the bruins yesterday so my answer for this one is none 😤😤” [(this was in reference to a pre-season game) lmao sis yall are okay. it was yalls babies against our roster players, I would have offed myself had the outcome been any different lmao]

Bonus

“Does anyone actually play for the red wings” [no]

“filip "thot" zadina” [don’t....]

Los Angeles Kings

The only right answer is Alec Martinez (41.56%--96 of 231 votes)

Adrian Kempe (38.10%--88 of 231 votes)

Anze Kopitar (12.12%--28 of 231 votes)

Other* (8.23%--19 of 231 votes)

*= Inconclusive results.

Write-ins

“jeff carter would snort a line of coke with gritty” [uhhhhh WHAT]

“uhh wayne gretzky...” [jvfluhddsf sis...]

“I couldn't name anyone on this team if you PAID ME” [fjldfdhfh god I wish that were me, sorry annie u know I joke....]

Philadelphia Flyers

Claude Giroux (44.78%--103 of 230 votes)

Travis Konecny (HAHAHAHAHA that’s my ugly hot gremlin) (24.78%--57 of 230 votes

Other* (22.17%--51 of 230 votes

Wayne Simmonds (8.26%--19 of 230 votes)

*= Nolan Patrick is apparently who y’all think is the 3rd hottest flyer, even tho he Looks Like That rn lmao. fuckin’ lettuce head.

Write-ins

“Gritty's googly eyes are the windows to the soul”

“andrea helfrich” [ur right]

“tk, because country boy i LOVE you 😛”

Bonus

“hey don't make threats abt gritty like that” [I put “if you put gritty i’ll block you”]

“My hellspawn son [Gritty,] is beautiful can’t believe Voracek and G had a son tho” [HDKUHEDKJFHD BITCH]

Pittsburgh Penguins :(

Kris Letang (55.17%--128 of 232 votes)

Other* (19.40%--45 of 232 votes)

Not Sidney Crosby [this is the option for Sidney Crosby] (16.81%--39 of 232 votes)

Tristan Jarry (8.62%--20 of 232 votes

*= different variations of “none” won but only by one vote, the person right behind was Jamie Oleksiak.

Write-ins

“the penguins roster came into my home and killed my entire family, but jamie oleksiak is 6'7" 255 lbs of A Man” [NDKFHSJRFDRBSKRFH valid]

“no penguin has ever been hot. As soon as they put on the jersey the hotness evaporates. Tragic.” [wow look at all that truth right there]

“as a heterosexual i chose letang, and as a flyers fan i choose the penguin mascot” [lmao girl letang is not the answer either]

St. Louis Blues

Colton Parayko (67.11%--151 of 225 votes)

Alex Pietrangelo (17.33%--39 of 225 votes)

Other* (8%--18 of 225 votes)

Ryan O’Reilly (7.56%--17 of 225 votes)

*= Inconclusive results.

Write-ins

“this [’other’] box shouldn’t exist there are no valid arguments against the angel colton parayko” [tru, but the blues have other hotties so I made the box to be fair to those of us who don’t like Big Blonde Sexies]

“uh valid i guess? idk any of the blues lmao” [LMAOOO I think they meant Vladdy, but “valid” cracked me up]

“ROR can lay me down” [ur so valid lmao]

Buffalo Sabres

Jeff Skinner (60.18%--136 of 226 votes)

Rasmus Ristolainen (17.26%--39 of 226 votes)

Other* (14.16%--32 of 226 votes)

Jack Eichel (8.41%--19 of 226 votes)

*= Inconclusive results. [Y’all big mad that I put Skinner on here. HE’S HOT!]

Write-ins

“Idk but not these lmao” [*instert that gif of the kardashians like “DON’T BE FUCKING RUDE”*]

“Why is Jeff Skinner an option he looks 12″ [who else tho sis. I looked at the roster!]

“If anyone says eichel i will come to their house and steal their toothbrushes. Its conor sheary.” [I took my own survey and picked Eichs but I still have my toothbrush so I guess......... I’m right.]

Bonus:

“Rasmus Ristolainen kinda looks like a creepy half-alive Ken doll, but I'll stand by my choice. Hire an exorcist.” [JDFKHRFWEH GIRL]

“They lost their only cute player when O’Reilly got traded sorry” [boom. roasted]

Vancouver Canucks

Brock Boeser (67.56%--152 of 225 votes)

Other* (13.78%--31 of 225 votes)

Jake Virtanen (12.44%--28 of 225 votes)

Ben Hutton (6.22%--14 of 225 votes)

*=Inconclusive results.

Write-ins

“[about Jake Virtanen] all that ass...........” [sjdkfhdkfhdkhfi yeah]

“the city of vancouver” [?????????????????????]

“I keep forgetting that the canucks actually exist” [I’m reasonably sure this is annie lmfao]

Bonus

“I don’t know how any of this team looks either” [idk if I follow Nucks blogs or what but how do u not know Boeser???]

“i don't care enough about this team to even attempt to answer” [this is my brain @ me on the last 5 questions of an exam]

New York Islanders

Mat Barzal (67.69%--155 of 229 votes)

Tito Beauvillier (14.85%--34 of 229 votes)

Jordan Eberle (10.48%--24 of 229 votes)

Other* (6.99%--16 of 229 votes)

*=Inconclusive results.

Write-ins

“you say put full names but then u go and say tito??” [LISTEN I was tired at this point and forgot that I was trying to be at least a little bit professional about my thirst survey alright? yeesh]

“Its Matt Martin my dude” [LMAO u funny]

“idk how anyone pays attention to mat when tito is always there looking better barzal looks like every attractive jock ive ever met and i dont trust that”

Calgary Flames

Noah Hanifin (37.95%--85 of 224 votes)

Elias Lindholm (32.59%--73 of 224 votes)

Matthew Tkachuk (20.54%--46 of 224 votes)

Other* (8.93%--20 of 224 votes)

*= Sean Monahan wins 4th hottest.

Write-ins

“[Hanifin] looks like the bad guy in a teen movie. the guy the Main Girl is dating in the beginning but is a real dick to her. you look at him and you KNOW he has a trust fund and votes republican. god he's so hot though” [hanny......... yeah.... yeah....]

“Why do I find Tkachuk attractive? I don't know but I love him” [me too]

“James 'The Real Deal' Neal” [lol I got this answer multiple times]

Washington Capitals

Tom Wilson (31.33%--73 of 233 votes)

Andre Burakovsky (29.18--68 of 233 votes)

Braden Holtby (24.03%--56 of 233 votes)

Other* (15.45%--36 of 233 votes)

*= Michal Kempny and Nicklas Backstrom tied for fourth hottest.

Write-ins

“literally no one, i s2g if i see anyone say ovi is attractive..... jfc god help them” [.... but ovi is dad-hot, also he got 3 votes]

“YOUR STANLEY CUP CHAMPIONS! Everyone btw just a hot team of hot ugly men and Tom Wilson” [kskdjskdjksks]

“my sweaty swedish sweetheart; Nicklas Backstrom” [I’m too illiterate to read this right the first time thru lol]

Colorado Avalanche

Gabe Landeskog ( 55.95%--127 of 227 votes)

Other* (22.47%--51 of 227 votes)

Erik “Horsegirl” Johnson (14.1%--32 of 227 votes)

Mikko Rantanen (7.49%--17 of 227 votes)

*= Tyson Barrie won by more than double of all the other write-ins, but honorable mentions go to Nate MacK, Colin Wilson, Tyson Jost, Phillip Grubauer, and The Avs Tumblr People.

Write-ins - I (jokingly) got called bitch so much in these write-ins, y’all feel some type of WAY about this team lmfao.

“but also the tysons. i would buy a whole farm just so those boys could plow me into the ground.” [i’m SCREECHING. this killed me lol]

“only attractive b/c of his proximity to horses? maybe so.” [.... girl what]

Okay, so instead of a third quote, cause I couldn’t pick, I’m gonna put all the other funny EJ comments I was contemplating:

“ej is soooo ugly in the hottest way possible”

“erik "big horny" johnson”

“oh my god Ej was included for once I'm weeping tears of joy”

“What that mouth do EJ?”

New Jersey Devils

Miles Wood (36.12%--82 of 227 votes)

Taylor Hall (33.48%--76 of 227 votes)

Brian Boyle (19.82%--45 of 227 votes)

Other* (10.57%--24 of 227 votes)

*= Nico Hischier with the majority of the write-ins, even tho he’s still a CHILD (under 20).

Write-ins, aka Mostly Taylor Hall Commentary.

“Does Michael McLeod count” [YES girl i love that boy]

“Gucciiiiiii”

“DSL GUCCI”

“Nico Hischier (Taylor Hall I still love you)”

“i chose taylor and i don’t even need a gucci purse”

“If Taylor Hall gave me a Gucci purse I'd vote for him”

“catch me w/ a gucci purse, girl!!!! for real tho miles wood”

Dallas Stars

DISCLAIMER: I mean no disrespect to Katie, she’s fab and I made this survey a month or so ago. If you don’t know what I mean by this--do not ask me, I will delete the message. Thank you!

Tyler Seguin (46.96%--108 of 230 votes)

Katie Hoaldridge (im gay) (35.22%--81 of 230 votes)

Other* (13.91%--32 of 230 votes)

Stephen Johns (3.91%--9 of 230 votes)

*= Jamie Benn.

Write-ins

“tyler seguin has no upper lip” [I screamed, not exaggerating]

“You have to choose [Seguin] but I do so under duress”

“Im gay too” [hell yeah, this is a mlm and wlw friendly survey!]

Edmonton Oilers

Jujhar Khaira (28.57%--64 of 224 votes)

Other* (27.68%--62 of 224 votes)

Contract McMoney (he is hot) (25.89%--58 of 224 votes)

Darnell Nurse (17.86%--40 of 224 votes)

*= Leon Draisaitl won by more than 5 times anyone elses write-in lmao.

Write-ins ft. “The Draisaitl Quotes”

“McMoney’s money- just his money” [lmao ok sammie, HE’S HOT!]

“cannot mcwingames went off in the gq shoot i admit” [*annie voice* OHMYGOD]

“He’s [Khaira] like a romance novel cover like, f me” [tru]

Drai Quotes

“Drai but like lucic cause Momma needs a man that could kill me” [HDGFDHDGFHDH]

“leon dreisetl (is that his name, is this how you spell it?)”

“Leon Draisaitl and his contract that he doesn't deserve” [backhanded compliment lmao]

“the one w the longass name. dry saitl or whatever” [girl. lmfao]

Winnipeg Jets

Jets/laine fans are funny so I’m adding all the funniest ones instead of just 3 or 5. Sorryyyyyy I’m here to entertain.

Blake Wheeler (44.04%--96 of 218 votes)

Mathieu Perreault (but specifically in his newest headshot) (21.56%--47 of 218 votes)

Other* (19.27%--42 of 218 votes)

Connor Hellebuyck (15.14%--33 of 218 votes)

*= Patrik Laine, even tho I said NOT TO, demons.

Write-ins

“Their logo so I can fly away from this stupid team”

“Nobody but I just needed to point out Connor Hellebuyck looks like a stage magician and that is Not Hot” [i respectfully disagree with the last bit but the first parts made me snort]

“I don't know who windy pegg is”

“Boeing 747″ [sjdjsljlshgdu]

“they’re all second to jacob trouba’s dog Donnie”

“Patty Laine, but like, without the demon beard”

“Let me live my life! Laine has a good voice and i have a LANGUAGE KINK!”

“Laine WITH the beard because I don't fear death”

“Sorry, Laine but only with his beard” [I love the halfhearted apology]

“Goatboi”

“ALL HAIL THE GOAT DEVIL”

“laine come at me bitch lol” [denny’s parking lot. 3 am. be there.]

“laine looks like a goat”

“Laine’s Beard”

“LAINE I like the beard but hockey Satan is good to hellebuyck” [I really felt like I was tripping balls while reading all these but, ESPECIALLY this one lmfao]

Arizona Coyotes

Oliver Ekman-Larsson (30.32%--67 of 221 votes)

Jakob Chychrun (28.05%--62 of 221 votes)

Dylan Strome (26.24%--58 of 221 votes)

Other* (15.38%--34 of 221 votes)

*= Alex Galchenyuk, with the majority of the votes.

Write-ins

“pls date me Chych” [annie, that’s my BF!]

“ 🐼 there is no raccoon emoji >:(”

“[Chychrun] [a]lso has a vampire quality but like trust fund baby vampire who has no morals. I’m...into it??” 

Honorable mentions: The 2 people who put Biz lmaoooo I love yall.

Carolina Hurricanes

Andrei Svechnikov [he’s a baby but I didn’t know who elseeee] (38.29%--85 of 222 votes)

Haydn Fleury (35.59%--79 of 222 votes)

Other* (15.77%--35 of 222 votes)

Dougie Hamilton (10.36%--23 of 222 votes)

*= Sebastian Aho wins the write-in vote [he ain’t it!]

Write-ins

“Justin Faulk (I’m old so svechnikov is out)” [ugh ur right I didn’t make this more inclusive to people not my age, i’m (genuinely) sorry!!!]

“Formerly Eric ‘the hottest Staal' Staal” [only on the cane’s write-in would I have this happen...]

“[Jordan] staal terrifies me but that's hot” [true!]

San Jose Sharks

Erik Karlsson (70.04%--159 of 227 votes)

Martin Jones (17.62%-- 40 of 227 votes)

Other* (11.01%--25 of 227 votes)

Justin Braun (idk) (1.32%--3 of 227 votes)

*= Inconclusive results.

Write-ins

“Daddy shark (doo doo doo)” [just so yall know this is, of course, annie, as in anzekopistar, an actual demon, she’s talking about Erik Karlsson :)]

“Brent Burns, you know im right” [are you tho?]

“Okay sometimes I have needs I think Joe Thorton sans beard could fill” [this is why joe shaved. he felt this person in the universe wanting him to, so he did, wow thank u joe]

Ottawa Senators (lol)

Matt Duchene (33.63%--75 of 223 votes)

The entire team (cause they’re a dumpster fire) (30.94%-- 69 [it’s that tkachuk fuckboi energy] of 223 votes)

Other (there are none)* (22.87%--51 of 223 votes)

Spartacat (12.56%--28 of 223 votes)

*= Inconclusive results (because a lot of you took my “there are none” joke a little too seriously and just chose that, no write-in lmao)

Write-ins

“[about Duchene] he's traitorous but it's like that sometimes i guess” [sjdhdjfhkdhf girl it’s okay.]

“.... we're a team“ [i-]

“the senator on their jerseys is p cute ig”

Bonus:

“oh so spartacat is an option but not gritty huh” [LISTEN the flyers are a HOT team, the sens are NOT. that’s why lmao]

“Just based on headshots I’m going with Ben Sexton like also how do you go wrong with that name”

Tampa Bay Lightning

Brayden Point (55.25%--121 of 219 votes)

Other* (22.83%--50 of 219 votes)

Mitchell Stephens (11.87%--26 of 219 votes)

Steve Yzerman (10.05%--22 of 219 votes)

*= Inconclusive results. Although there were a lot of responses none of them added up significantly sooo....

 Write-ins

“am i the only one who thinks stevie y was a bit of a twink when he was younger?” [jdhslihdalskdjefh]

“Worst team in the league i hate them and theyre all hideous” [u sure bout that, bud?]

“Stamkos (I love his tiny eyes)” [????]

Florida Panthers

 Aaron Ekblad (71.75%--160 of 223 votes)

Aleksander Barkov [r yall ok???] (11.66%--26 of 223 votes)

Other* (10.31%--23 of 223 votes)

Vincent Trocheck (6.28%--14 of 223 votes)

*= Inconclusive results.

Write-ins

“who are the panthers” [sometimes a team is a dog captain, a(n extremely hot) 27-year-old lawyer, and not owen tippett because the panthers hate me specifically]

“Roberto Luongo during Parkland speech” [... valid]

“barkov is literally the only player i know on this team” [shey would be happy to teach u about the panthers!]

Anaheim Ducks

Adam Henrique (52.47%--117 of 223 votes)

No one else (29.6%--66 of 223 votes)

Other* (10.76%--24 of 223 votes)

John Gibson (7.17%--16 of 223 votes)

*=Inconclusive results.

Write-ins-Ducks fans don’t @ me but i’m pretty sure half of these were submitted by y’all anyways....

“if i look @ anyone on the ducks roster for more than 5 seconds i BLACK OUT” [KSHDGJDHSKH Adam tho....]

“Quack Quack go lay your eggs somewhere else you feathered FUCKS” [sjdjfhdjsksj]

“legal 2 say kesler?” [no. go to jail]

Bonus

“Henrique is fine I have no qualms about your selections” [thnk u]

“jared coreau!!! GOOGLE HIM i’m right” [I said this, and we’ve talked, but I need people to know that I, after seeing this, subsequently found out that the Wings didn’t sign coreau back this offszn lmao] 

Nashville Predators

Roman Josi (39.39%--91 of 231 votes)

PK Subban (37.66%--87 of 231 votes)

Kevin Fiala (13.42%--31 of 231 votes)

Other* (9.52%--22 of 231 votes)

*= Pekka Rinne for 4th hottest. [My mom loves him for his name lol. she says it’s “fun”]

Write-ins 

“pk wears cool hats. I like that in a man”

“I don't find any of them hot (Josi used too be hot and then I learned he was illiterate and now I feel nothing but pity towards him)” [GIRL]

“preds are also ugly. pk subban would be attractive if he werent a pred” [lmao. what’d they do to u ?]

Columbus Blue Jackets

Pierre-Luc Dubois (50.22%--113 of 225 votes)

Zach Werenski (20.44%--46 of 225 votes)

Josh Anderson (16.89%--38 of 225 votes)

Other* (12.44%--28 of 225 votes)

*= Alex Wennberg is 4th hottest [lmao]

Write-ins

“Can I put werenskie and Anthony Duclair” [valid]

“Just to be clear CBJ is by far the hottest team exemplified by the fact that you left Seth Jones and Alexander Wenneberg off this list when they're like top 20 in hotness. Also Nick Foligino wins if we include looks and personality.” [I didn’t include them cause this is a mix of hot and ugly hot fam, the avs are 100% the hottest team in the NHL, and that’s coming from me, a Wings fan, destined to hate the Avs for my entire life. Also???? The hotter Foligno is def Marcus lmao]

“[About Werenski] only with the scar though otherwise seth jones” [GIRL scars don’t disappear??? WDYM only with the scar??? Are you a time traveler??? lmfaooo]

Minnesota Wild

J.T. Brown (46.46%--105 of 226 votes)

Other* (21.68%--49 of 226 votes)

Eric Staal (20.80%--47 of 226 votes)

Jason Zucker (11.06%--25 of 226 votes)

*= Charlie Coyle. Honorable mentions to Zach Parise and Matt Dumba.

Write-ins

“Charlie Coyle man! V hot, could kill you, gently waves at babies, 10/10″ [exactly my type! wow]

“ Not JT[,] Lexi is the hottest[,] Eric Staal from a few years ago is also hot” [I added commas to your thing cause.... girl it took me a sec to understand what u were tryna say. But also ur right it’s Lexi.]

“love a #wokebae jt” [yaaas]

FINALLY this legit took me like 10+ hours of work cause I had to transcribe all the info cohesively and then go thru all the responses lmao.

Vegas Golden Knights

William Karlsson (40.52%--94 of 232 votes)

Marc-Andre Fleury (30.60%--71 of 232 votes)

Max Pacioretty (16.81%--39 of 232 votes)

Other* (12.07%--28 of 232 votes)

*= Inconclusive results.

Write-ins

“[Karlsson] because he looks like young Bill from Mamma Mia” [shfhdjdhf girl]

“fleury isn't hot you absolute monster” [???????? drink ur bitterness tea somewhere else pls]

“let's find out just how wild this boy is” [pftd dtduftdhjfgdjfghdjf]

Bonus/Honorable mentions:

The TWO people who put “colin miller’s eyelashes” lmfaoooo

Alrighty this is The End! If you’d like to see another survey by me let me know in my messages/ask!!! Also sorry for stealing de la Rose from u, habs fans

[tigerlily-sunshine]

Can we expect anymore surprises? Cause I think calum done been hurt romantically before by some1 bc in the 1st chapter they mention that luke done brought some1 home before

First off, beware that this answer might be more spoiler-y than I usually do. 

I think there is a misunderstanding. Perhaps I wasn’t clear enough in the story or maybe it is a conversation that sounds like one thing but is actually something else. (There is a lot of instances of the latter purposefully woven throughout the story.) To be honest, I am not sure which part you are referring to in the beginning of the story where there is any mention that Luke has brought someone home before. That didn’t happen. 

(Feel free to send me the exact passage, and I’ll either correct my error - because it is totally possible that I misstated something in the narrative/dialogue - or I’ll explain what it actually does refer to.)

What I think you may be referring to is the scene in chapter four where Ashton realizes that Luke and Calum and Michael are in a relationship together. Within that whole chain of events, there is a conversation between Michael and Luke about Luke “always doing this” and “always getting his heart broken” and “never wanting to leave Michael and Calum behind.” (These aren’t exact quotes but rather Michael’s three main points.) That conversation will make a whole lot more sense when Luke’s part of their history is revealed. I don’t want to trample all over the suspense of that future scene (and by future, I mean it is the chapter after the next one that will be posted). 

There is a reason that Michael hammers Luke for not telling Ashton about the exact nature of Luke’s relationship with Calum and Michael. It isn’t because he has brought someone home before. It is more because of what Michael and Calum’s relationship has meant for Luke all this time. (Some reference to this is found in chapter six when Calum “explains” how the relationship between him, Michael, and Luke began. But be wary of that scene for reasons.) 

But to answer your question: are there any more surprises before the end of the story? Yes and no. There isn’t an unmentioned past, scorned lover lurking in the shadows. All of the characters in play have been established long before now.  

Should there be surprises? …No, not really. Tiny details, maybe, like why Calum drives such an old car or what exactly Brandon is up to or why Michael even fell in love with Luke in the first place. Every thing else? All of these secrets? If you read story close enough - if you pick apart every single scene with the understanding that everything is distorted and purposefully misplaced and lineally misleading - the ending will not be that surprising. 

But the truth is that basically nobody is going to read this story as closely as that, at least not without knowing what all of these truths are. And no one is going to pick apart the truths from the many, many not-quite-truths. So I am being a little facetious by saying there are no surprises. There are. In fact, my hands-down favorite scene in the entire story is absolutely unpredictable. (But hilarious and, trust me, it works out better than any other alternative. That scene is far off, nearly at the end, though.)

What I think you’re asking is simply whether readers should expect something out of the blue? An ex-lover or a wrench in the plans that the story hasn’t even touched upon? That answer is no. What is left of the story is tying up loose ends, revealing the whole truth behind everybody’s secrets, and paving the way for a happy ending where the four of them are together. Oh, and, you know, dealing with that whole Brandon thing. ;)

(Of course, feel free to correct my assumption of your question or to ask follow up questions if you are still curious.)

I seriously considered leaving a snippet of Luke’s part of the story, but that in itself is very, very spoiler-esque for reasons, so I’m going to steer clear of that and leave this from the next chapter instead.

There are only a coupleof bags left for him, and they are waiting on his unmade bed. Ashton tries notto think about how Luke would sneak in here every night no matter how lateAshton arrived home just so they wouldn’t have to sleep alone. He can’t focusthings like that, just like he can’t think about how Calum took to waking himup and fixing him a mug of oatmeal for breakfast after long nights spent at theoffice. If he thinks of the niceties of believing that Calum and Luke andMichael wanted Ashton as much as he wanted them, he is liable to break downcrying and never leave.

[drugoftime-blog]

This post leads on from last week’s on antagonists, but is aimed a slightly more specific type of antagonist - the villain. In this, I’ll be looking at some of the most famous antagonists from fantasy-adventure works - the ones who genuinely strike fear into the hearts of our protagonists and are a real threat; the ones we remember long after the story is finished. Villains like Lord Voldemort, Darth Vader, The Witch-King of Agmar, and even the White Witch from The Lion, the Witch and the Wardrobe.

It should be noted that all of the suggestions I give in the post come in tandem with last week’s post. You should not forget as you read that your villains, like all antagonists, need real motivations, arcs and faces to be interesting.

That being said, how can you write a fantasy villain that inspires that deep sense of fear, power and gravitas?

What Not To Do

Tell, rather than show.

Please don’t write: Lord EvilMan, the most powerful and scary man in all of EvilTown, walked in the room. He looks at the people and says: “I’m so powerful. I could kill every single one of you. Don’t cross me!”

Okay, my example is a bit exaggerated, but the point stands. There is nothing less impressive than a villain telling you how powerful he is, or you simply telling the reader that he is very powerful and scary. More than most things, these traits need to be shown to be effective.

So how do you show these things? Well, here are some ways to consider it:

Reactions and Fear

One of the things that make all the above villains so ominous is the way that other characters react to them over the course of the story. When we discover who fears them, we discover a lot about how dangerous the character is.

Sometimes, it will be the way they react to their presence, for example when Darth Vader enters a room, people will cower and look subservient without him really needing to say anything. Film, of course, has the added benefit of being able to throw some great music and sound design onto this as well, but what is important is that you are capturing the tone in your writing. You should be trying to capture this in your prose and one effective way of doing then is describing how people - especially strong and important people - react to them.

This is even more impactful when the villain isn’t there. Consider how we are introduced to the White Witch in CS Lewis’ wonderful The Lion, The Witch and the Wardrobe. The beavers are extremely scared when her name is mentioned. Tumnus is so scared that he actually betrays his new friends, just because he is worried what will happen to him if he doesn’t. By making the characters do things that are out of character and against what they would usually do at the mere mention of the villain, without the villain actually being there, tells us a lot about their power.

The most obvious example of this working is Lord Voldemort. The fact that before we even meet him people are too scared to say his name in public has a huge effect on our perception of him as a reader. It’s a very clever narrative trick.

This is amplified by the fact that even after he has been defeated people are still too scared to speak his name. They actually flinch when others do it. This helps paint him with all the danger and gravitas that he deserves and means that when we do actually see him rise again at the end of The Goblet of Fire, it becomes one of the most memorable scenes in modern literature. This is because of how Rowling has set him up using others reactions.

I’m not saying that everyone needs to be scared to say your villain’s name. That would be a bit too on the nose and bit too similar to Rowling, I think. But do consider other ways that you can have common and normal people react to discussing or mentioning the villain. This really helps build up their danger.

While having common folk fear the villain is a good thing, sometimes this is even better demonstrated by someone very powerful fearing the villain. In the Lord of the Rings, Gandalf is set up as being an extremely powerful wizard, so when he becomes afraid of the Balrog, we know things are getting serious. This happens again later on where Gandalf is used as a narrative tool to indicate that the Witch-King of Agmar is someone to be really scared of, as he has no desire whatsoever to fight him.

If you can have otherwise powerful and competent heroes suddenly become quite afraid of something or someone in particular, this contrast will really help develop the villain’s sense of strength and danger.

Stories and Mythology

Real worlds and cultures have stories and myths built into everything. Don’t forget about this when you build your world. In The Name of the Wind, the Chandrian appear in fairy tales and songs sung by children. They are very much an old part of the world and that lends them a certain gravitas.

Imagine if your character is walking through a town and their one claim to fame is, “one day Lord EvilMan stopped here to stay the night. Up there is the house he stayed in, they say no one can stay a night in there without going mad.”

This is a demonstration of how your villain becomes part of the cultural tapestry of your world. Name things after them. Have stories, apocryphal or true, told about them. Make people scared.

Real Power and Threat

Your villain needs to pose a real threat to your protagonist. A real threat. There are two things you need for this to be the case.

1. Have them actually do something

Please don’t make your villain all talk and no action. Good examples:

The Chandrian murder Kvothe’s entire family at the start of The Name of the Wind.

Darth Vader orders the destruction of a whole planet. Then he fights Obi Wan and kills him. Then chops off Luke’s hand and throws him down a shaft.

Give your villain some real gravitas and don’t wait until the second half of your story to do it. If you want us to be intimidated and threatened by them, give us reason to be.

2. Imminent Threat

Great villains work well if the protagonist feels like they could be killed at any time. The threat feels very imminent. For all that I love the books, and the villains, my mine gripe with Rothfuss’s Kingkiller Chronicles is the fact that the Chandrian appear so violently at the start and then basically disappear for the rest of the books. They stop feeling like a threat and lose all of that great fear that was built up at the start.

There are many ways of creating an imminent threat. The White Witch is a constant and imminent threat because she has Edmund captured and is slowly turning him against his siblings.

3. Have them far more powerful than your protagonist

Nobody wants a villain who is easily defeatable. The best protagonist / villain relationships are ones that are hugely unbalanced. Our villain needs to be a real threat. Our protagonist should be the severe underdog in any conflict.

Consider Lord Voldemort - apart from some other clever narrative tricks like having the same wand, Harry doesn’t stand a chance. If it was a normal magical duel with no extenuating we all know that he would be dead. There’s no question of the matter.

What power does Frodo have to defeat the Witch-King of Agmar or Sauron for that matter? None, except for the possibility of destroying the ring. But in one on one combat they are a clear and impossible threat.

The White Witch even kills Aslan, our great hero and saviour, without much of a fight. (I know, I know, he comes back, but at the time it has a lot of impact).

A great example of this actually comes from Brandon Sanderson’s YA novel, Steelheart. I think Steelheart is a great villain precisely because he is so completely unstoppable. When they do stop him, it isn’t because the protagonist is stronger or more powerful, but because of another narrative twist.

Conclusion

If you want a really scary, memorable and intimidating villain, consider doing the following:

Show how scary they are by using reactions of other characters, both big and small.

Make them part of the cultural and narrative tapestry of your world.

Make sure they act early on, and don’t just talk about it.

Make them an imminent and real threat and make them far more powerful than your protagonist.

Discussion

This post is designed to both help inspire, but also be a springboard for discussion.

What do you think makes a great villain?

Do you disagree with any of the above?

What are your favourite villains in stories and what makes them work? Do you have other good examples for us to pick apart?

*

(without wanting to make this about self-promotion, if you found any of this interesting, feel free to check out www.binge-writing.com - there will be updates every Thursday.)

As much as I strive to post once a week, next week I will be away in Sri Lanka, so may not be able to.

If I am unable, then the week after I will be writing about how to construct individual scenes and chapters and what goes in them.

See you in two weeks, probably.

[tran5rightsos]

You’ve Cut the Wrong Damn Wire - Chapter One

Tags and Warnings: Implied/Referenced Prostitution, Character Death, Blood and Gore

Word Count: 1389

Leave Kudos?

Calum knew him long before he actually met him. Everyone knew the skinny blond with the big, sad doe-eyes, by reputation more than anything else. Luke Hemmings had more than his share of rumours and gossip going around.

Maybe it was the people he hung out with, but Calum hadn’t known many people who could actually confirm any of it. It was always, “I heard from Brandon that he’ll give you a handy for ten dollars and a blowjob for twenty,” and, “Jack says it’s more than that, at least thirty for the handjob,” and, “Trent reckons he’ll do all of it and more for free if you give him a bit of xan every now and then.” The nicer stories were that his parents were dead or in prison or dumped him on some relative when he was born.

Calum couldn’t say that he was inclined to believe many of them, it was a pretty well-known fact that Luke was an outcast, a weirdo. He was easy to pick on because he was small and quiet, so a lot of people did, safe in the knowledge that he never went to anyone about it and nobody cared because, “He’s the kid that’s always high first period.” Calum had seen him walking to classes late with some seriously nasty-looking bruises plenty of times, hunched over and looking like he wished nobody could see him.

Although he didn’t know Luke very well, Calum was glad when Michael Clifford started hanging out with him in year twelve. Michael had rumours of his own, but they were more along the lines of, “He stabbed a kid in primary school,” and, “He killed his neighbours’ cat,” and, “He got expelled from his old school for beating people up.” From what Calum knew of him personally, Michael was perfectly friendly. His bright hair and lipring might’ve been off-putting to nervous old people, but he smiled and laughed a lot and the most Calum ever saw him get in trouble for was dress code violations, though teachers did seem to call him out on them at the beginning of every class Calum shared with him. It didn’t matter. Just as with Luke, the rumours were there and they weren’t going away.

Year twelve was when Calum noticed Luke hanging out outside of school for the first time, too. It had been weird to see him out of uniform and somehow Calum had never imagined that he was the type to eat ice cream outside a club, laughing about something with Michael. It felt good to see him happy. A few people wondered if they were dating, but everyone knew that you’d have to be a real asshole to have a problem with that.

Whatever the truth of everything was, while people still sometimes talked shit about Luke Hemmings behind his back, nobody wanted to fuck with Michael Clifford. When Luke showed up to school limping and sporting a black eye, nobody bragged about it being them anymore.

After graduation, Calum stopped seeing Luke around as much, though Michael often came by Zig Zag’s, the ice cream place he was working at, for a cone each of rocky road and strawberry. He even learned Calum’s name and started greeting him whenever he came in, a friendly grin on his face. They never spoke much, but Calum got the impression that he and Luke were happy together, if they were together at all. A pretty coworker of Calum’s, Ashton, once commented that he rarely saw Michael or Luke alone, that they were almost always near one another, but Michael liked to flirt so it was hard to tell for sure. Calum had been tempted to give him his number a few times, but he worried that it would hurt his chances with Ashton.

Ashton Irwin had been working at Zig Zag’s for a few years and took it upon himself to train Calum when he was hired. He had a handsome face: green, hawk-like eyes, slightly pouty lips and a pointed chin and he always looked out for Calum when he fucked up or needed help with something. He was only a few years older than him, but Calum found the age difference intimidating and had a hard time working up the guts to ask if he wanted to hang out outside of work sometime.

Luckily for him, Ashton was the one to ask him if he wanted to go to a small party at a friend’s place one weekend. Calum was keen for both the opportunity to get to know Ashton better and to catch up with some high school friends he hadn’t seen in a while, so he was quick to give him a shrug and a, "Yeah, why not?"

Ashton picked Calum up on the night of the party, dressed in a leather jacket and black skinny jeans, his hair gelled back aside from a few wayward strands. He looked like a bad boy from a 70s movie and Calum loved it.

At the party, Calum didn’t drink much, worried that he’d do something dumb and embarrass himself in front of Ashton. A few friends from high school said hi and he took the opportunity to act cool and totally not needy or clingy, chatting about what everyone had been doing since graduating and what their plans were, even though he’d have liked nothing more than to listen to and stare at Ashton all night.

It wasn’t a date, but Calum managed to enjoy the night all the same, leaving late with Ashton and wondering if he could still spin things in a date-like direction. As they walked back to Ashton’s car he mulled over whether he could ask if he could spend the night at his place, or if he should play it cool and simply suggest that he’d like to hang out with Ashton again sometime.

Before he could come up with anything, Calum spotted something in the ditch that separated the road from a park.

“What’s that?” he wondered, pointing.

Ashton peered at it. “Looks big. Kangaroo, maybe?”

As they got closer, Calum turned on his phone torch to shine it on the mystery thing.

He froze. “Holy shit.”

The body curled up in the ditch was a mess. As Ashton called 000, Calum stared at it, at the wet blood and tattered clothes, at the blond curls and lanky limbs, at the horribly disfigured face and the glassy blue eye that stared at nothing. His stomach twisted.

“Luke?” He didn’t know who he was asking, but he hoped that they would tell him that no, it wasn’t Luke Hemmings, the lonely boy from school who was always so quiet, who always kept to himself and never gave anyone shit even if they deserved it.

“Hey,” Ashton said softly, wrapping an arm around his shoulders and taking his phone to turn off the torch.

Calum hadn’t realised that he was crying. He wasn’t even sure why he was crying, he barely knew Luke, had never even had a full conversation with him. He was crying, though, and he kept crying as the police pulled up and Ashton told him not to say anything, he’d sort it out. He kept crying as people from the party realised something was up and gathered around, whispering and trying to get a good look at the small, broken remains. He kept crying as they got in Ashton’s car and drove off, his mind whirling with how sick the world was to give Luke such a cruel life and then end it so horribly.

Ashton took him back to his place because it was closest, a small apartment in a quiet area, and held him as he broke down. The next morning, Calum blamed his tears on the alcohol.

[aion-rsa]

Best DC Comics to Binge Read on DC Universe

https://ift.tt/2LjeV6J

With an enormous swath of the world involved in varying degrees of social distancing, many of us suddenly find ourselves with a lot of time on our hands. Never fear! There are more options for streaming comics than ever before, and that means we have access to more of comics history, more hidden gems and epochal runs than ever before. But the variety of options to read can be daunting. That’s why we’ve put together a recommendation list of some of our favorite comics binge reads to help you through quarantine.

DC Universe rolled out in 2017 as the first full-service entertainment streaming platform – old shows, old movies, new shows, new movies, and a huge library of comics. And while a lot of the excitement over the platform has been about that original or new shows (justifiably! Harley Quinn and Doom Patrol are amazing!), it also gave us access to a staggering catalog of old comic books. 

If you’re coming to a comic streaming service like DC Universe, chances are you don’t need us to recommend the hits. Nobody who watches the CW shows needs to be told that Crisis on Infinite Earths is worth reading. Likewise Batman: Year One, or All-Star Superman or The Great Darkness Saga. We’re going to skip over some of the obvious ones and point you towards hidden gems, stories you might have otherwise skipped over but for a trusted recommendation. We are also looking for monster runs that will keep you occupied – you can read six issues in one sitting. Some of these might take you an entire round of social distancing to finish. 

A quick note about the reading guides: Many of them may have their own separate entry under DC Universe’s reading lists – those are helpful, but these are definitive. We will occasionally link to non-Den sources, but if you like what you hear, you should be encouraged to find your own best path. A lot of these stories wend through crossovers that are of varying degrees of relevance to the main books. It’s your call if you want to read the whole thing.

The Death and Return of Superman

The Death of Superman Reading Order

I know I said we wouldn’t talk about obvious must reads, but I feel like The Death of Superman (and it’s aftermath, World Without a Superman, Reign of the Supermen, and Kal-El’s inevitable return) should be on here. They can’t really be recommended enough. 

“The ‘90s” are often maligned as a wave of gimmicks and stunts, and killing the most important comic character in the history of superhero books definitely qualifies as a stunt. But what made The Death of Superman stand out (and several other ‘90s DC events, to be honest) is that it was actually very good. This era of Superman comics is actually a hidden gem – Clark is a joy, and all the weirdness and fun of the Superman universe is in full swing, like Cadmus, Mxyzptlk, and a truly bizarre (but surprisingly good) Justice League roster.

Read more

Movies

Men of Steel: 11 Actors Who Have Played Superman

By Mike Cecchini

TV

How Brandon Routh Returned as Superman for Crisis on Infinite Earths

By Mike Cecchini

The four writers – Jerry Ordway, Louise Simonson, Roger Stern, and Dan Jurgens – move pretty seamlessly between them on the main Superman books, and the art teams (Jon Bogdanove, Jurgens, Butch Guice, and Tom Grummett especially in the Death story) do amazing jobs of telling the story. Don’t be fooled by how gimmicky this feels, The Death and Return of Superman actually lives up to the hype.

Batman & Robin

Batman & Robin #1-17, Annual #1, Batman #17, Batman & Robin #18-32, Robin Rises: Omega, Batman & Robin #33-37, Robin Rises: Alpha #1, Batman & Robin #38-40, Annual #3

The Pete Tomasi/Patrick Gleason run on Batman and Robin never got the love it should have, because it ran parallel to two of the most high-profile Bat-comics of all time in Scott Snyder and Greg Capullo’s Batman, and the back half of Grant Morrison’s story in Batman Incorporated. But in ten years, people are going to be looking back at this as a classic. 

Read more

Comics

True Detective Creator Outlines What His Version of Batman Would Be Like

By John Saavedra

Movies

The Batman: Release Date, Cast, Villains, and More Details About the DCEU Movie

By Rosie Fletcher and 2 others

This is a controversial claim, but if you read this run, I think it holds up: Pete Tomasi writes the best Damian Wayne. He’s the right mix of arrogant little shit and not-actually-as-competent-as-Batman, and he actually learns lessons in this run that feel earned. He also dies during these stories, and Tomasi gets the chance to explore Bruce’s way of grieving, as well as drop in a series of guest stars that includes the best Two Face story I’ve ever read. Gleason and inker Mick Gray are utterly incredible, and do as much with one sixth-page panels with heavy inks and silhouettes as many art teams do with full page splashes. It’s a great, underrated run that I think you’ll love.

Wonder Woman

Wonder Woman (2006) #14-44, one story in #600

Oh my goodness Gail Simone’s Wonder Woman is exactly, precisely what I want out of a Wonder Woman comic. To me, Diana’s comics are an exception in that they should be as focused on how to avoid fighting as they are on the action. This run does that perfectly: she isn’t a belligerent meathead looking to stab everything in sight (but she does spend a little time with a neat Conan analogue, while we’re on the subject). She’s truly an agent of peace who then periodically has to kick some ass.

Read more

Movies

Wonder Woman Wasn’t Always Set During World War I

By Kayti Burt

Movies

Wonder Woman 1984: Who Is Maxwell Lord?

By Jim Dandy

The art is really good – Aaron Lopresti and Bernard Chang handle the bulk of it, and the storytelling and pacing are really well handled, but the panel borders stand out as especially interesting and visually entertaining. The guest stars are great – Black Canary brings Diana to Roulette’s fight club for a couple of issues, and there’s a big Power Girl punchup later in the run. This is just excellent, excellent Wonder Woman storytelling.

Suicide Squad

Suicide Squad on Comic Book Herald (end at issue #66)

John Ostrander, Kim Yale, and (mostly) Luke McDonough’s original Suicide Squad is a revelation. The concept is almost overdone at this point, and is a little bit ruined by putting big names like Harley Quinn on the team, but taking a batch of nobody villains and putting them on suicide missions to earn their freedom actually sets serious stakes, and this book does everything it should with those stakes. This is politics and espionage and force projection all wrapped into a story that makes the DC Universe feel more complete. 

Read more

Movies

Suicide Squad 2 Cast, Release Date, News, Story, and More Details

By Mike Cecchini

Movies

The Many Deaths of The Suicide Squad

By Marc Buxton

Beyond the plotting, though, there are so many great characters that come out of these books. Amanda Waller is one of the single best characters in all of DC Comics, and this is the run that made her the badass who can face down Batman in the shower without flinching. Punch and Jewlee are hilarious running gags. Deadshot gets some incredible work. Hell, even Captain Boomerang gets multiple dimensions added to him (without ever losing his core concept: he’s a giant asshole). I promise you, I’m underselling how good this era of Suicide Squad is.

Legion  of Super-Heroes

Legion of Super-Heroes Secret Files & Origins #2; Legion of Super-Heroes (1989) #122-125 alternating issues with Legionnaires (1993) #79-81; Legion Lost (2000) #1-12; Legion Worlds (2001) #1-5, The Legion (2001) #1-26, Legion Secret Files & Origins 3003; The Legion #27-33

If you loved Dan Abnett and Andy Lanning’s Marvel space work, when you read their Legion of Super-Heroes, you’ll be baffled at how Guardians of the Galaxy ended up on the big screen and not this. 

The Legion of Super-Heroes is generally regarded as…not the most newbie-friendly superhero team in the world. Fair or not, this run of Legion comics is incredibly accessible and does as good a job integrating them into the larger DC Universe as any I’ve read. It’s also exactly like DnA’s Marvel cosmic work, in that it is wonderful space opera that happens to have superheroes. The first batch of stories deals with a wave of catastrophes hitting the galaxy in quick succession. Legion Lost has a group of Legionnaires get thrown outside of the galaxy as they’re trying to fix one of the first catastrophes. Legion Worlds serves as a series of check-ins with popular Legionnaires left behind in the United Planets and is a really effective way to hook you into the 31st century of the DC Universe.

And finally, The Legion is an outstanding team book following all of those. Legion Lost is an unquestionable highlight; Olivier Coipel’s art is incredible, and the story will make you launch your tablet/phone/computer across the room at a couple of twists. This run is incredible comics. 

Justice League International

…you don’t have to read all of this, but if you feel like going for it, do it. You can stop at the red dots, though.

The Bwa-Ha-Ha era is half-superhero comic, half-workplace comedy, the template for greatness to come in Legends of Tomorrow, but a great superhero work in its own right. It’s an era of Justice League that takes itself (and its villains, and its stakes) much less seriously than just about any other era of the last 40 years. If you were raised on the post-Morrison “New Olympus” era of the League, the tone shift might be a little jarring. But that tone shift is part of what makes Keith Giffen, J. M DeMatteis, and Kevin Maguire’s run on Justice League special.

There are so many really good characters in this book, but one of the best parts is how much it does for both the League staples like Martian Manhunter and Batman, alongside the…less substantial…characters. Blue and Gold (Beetle and Booster, respectively) got their start here, and that one panel where Batman knocks out Guy Gardner that gets shared around the internet once a year is from this era.

Read more

Comics

Justice League Keeps Building the Wider DC Universe

By Mike Cecchini

Comics

New DC Universe Timeline Revealed

By Mike Cecchini

And besides being great comics, this run is also the favorite Justice League of a disproportionate amount of current comics writers, giving it an outsized influence on not just current books, but the rest of pop culture that superheroes have taken over – Wonder Woman 1984 is probably going to owe a HUGE debt to the Max Lord created by Giffen, DeMatteis, and Maguire.

Deathstroke

Deathstroke: Rebirth #1; Deathstroke (2016) #1-18; Titans (2016) #11; Teen Titans (2016) #8, Deathstroke #19-20, Teen Titans Annual #1, Deathstroke #21-42 (and when they go up, read The Lazarus Contract crossover and through issue #50 of the main series)

Priest’s Deathstroke is the best book that came out of DC Rebirth. Under normal circumstances, Slade Wilson sucks. He too often falls into a murder daddy archetype, a super cool anti-hero who goes big on the violence and the dysfunction as background statuses, and not as relevant parts of his story. Priest turned all that on its head and turned in a 50 issue run (plus a couple of specials, annuals and crossovers) that was about a father who loved his kids and didn’t know how to tell them, who also happened to be a top shelf mercenary and supervillain. 

Read more

Movies

Deathstroke Solo Movie Details Revealed by Gareth Evans

By Kirsten Howard

TV

Deathstroke: The Most Versatile Villain in the DC Universe

By Marc Buxton

That’s not to say there isn’t some super cool ass-whipping in it. Batman and Damian Wayne are recurring characters, as Priest sets up a mystery that might undo Damian as a character and gives more depth to Deathstroke’s issues with the Teen Titans. There’s an entire arc dedicated to him fighting various aspects of his own personality, personified in other villains from the rest of the DCU.

And it’s all so clearly and aggressively Priest – it has all the same style as his iconic Black Panther run, but with different storytelling to fit Slade’s tale. This is one of my favorite comics from recent years. 

Starman

Starman Reading Order on ComicsBackIssues

For about three quarters of my entire life, DC had an absolute stranglehold on legacy in superhero comics. The entire DC Universe was littered with stories about someone new picking up an old cowl and an old title and having to grow into that role, whether it’s Jason Todd as Robin, Wally West as Flash, Dick Grayson as Batman, Kyle Rayner, Connor Hawke, Tim Drake, Stephanie Brown. The list is nearly endless. The thing is, it’s a really good story archetype and an excellent use of shared universe superhero trappings to give heft and depth to stories that are otherwise not really allowed growth. 

Read more

TV

DC’s Stargirl Reveals Justice Society of America and Villains

By Mike Cecchini

Comics

Inside the Return of the Justice Society of America to the DC Universe

By Mike Cecchini

No comics did it better than James Robinson and Tony Harris’ Starman. It tells us the story of Jack Knight, the extremely Gen X son of golden age Starman Ted Knight. Ted is retired and passed his cosmic rod onto his son David, who gets murdered at the end of the first issue. It’s a hit on Ted’s whole family by one of his old villains, and Jack has to take up the rod to survive. Then he gets thrown into the mythology of the DC universe explained through the Starman legacy. It’s beautiful, fun, sad, meaningful, and heartfelt, and I bet you $1 that you cry at least once. 

The Question

The Question (1986) #1-15, Detective Comics Annual 1988 , Green Arrow Annual 1988 , The Question Annual #1, The Question #16-24, Annual #2, #25-36

Read more

Comics

The Question Bounces Through Time In New DC Series

By Jim Dandy

Everyone jokes about how much of scenic Gotham City is abandoned amusement parks and chemical plants, but Gotham City is a family-friendly resort compared to the Hub City of Dennis O’Neill and Denys Cowan’s The Question. “Atmospheric” doesn’t even begin to describe this run.

It takes The Question, a character created by Steve Ditko, co-opted and pastiched as Rorschach by Alan More and Dave Gibbons in Watchmen, and introduced him to the DC Universe proper by putting Vic Sage through a spiritual ringer. Everything about this book is incredible – Vic is a terrific character; his supporting cast is thoroughly real; the book ties into the greater DC Universe really well (via Richard Dragon, Lady Shiva, and the annual crossover in the middle with Batman and Green Arrow).

But the real star here is Hub City, a love letter that’s also hate mail to mid-80s urban blight as scenery. And Cowan and inker Malcolm Jones III’s art – it’s tremendous.

Orion

Orion (2000) #1-25

I’ve been a fan of Walt Simonson’s Thor since I first read it, because it’s obviously incredible. But I didn’t realize until Thor: Ragnarok and DC Universe came out that Simonson might be the best comic creator to follow up on Jack Kirby’s ideas of all time, and it was Orion that really did it for me.

cnx.cmd.push(function() { cnx({ playerId: "106e33c0-3911-473c-b599-b1426db57530", }).render("0270c398a82f44f49c23c16122516796"); });

Simonson puts Orion, son of Darkseid raised on New Genesis by Highfather as part of the peace treaty between the two factions of New Gods, on his prophesied track to kill Darkseid, and finishes it pretty early on. The fifth issue is just Simonson drawing a huge blowout fight between the two, and it’s predictably gorgeous. But he sticks with the story past that battle and digs deep into Orion’s character, the mythology of the New Gods, and some of Kirby’s best creations (the Newsboy Legion has a running subplot and it’s awesome). It also has backups from some of the biggest superstars in comics (Frank Miller and Dave Gibbons, among others). This is a hefty run of comics, but you won’t be able to put it down.

The post Best DC Comics to Binge Read on DC Universe appeared first on Den of Geek.

from Comics – Den of Geek https://ift.tt/35DlxWB

[fathersonholygore]

  Syfy’s Channel Zero Season 3, Episode 2: “Father Time” Directed by Arkasha Stevenson Written by Harley Peyton, Mallory Westfall, & Angel Varak-Iglar

* For a recap & review of the Season 3 Butcher’s Block premiere, “Insidious Onset” – click here * For a recap & review of the next episode, “All You Ghost Mice” – click here Open on a Peach’s Meat produced commercial called “Top of the Food Chain.” Mom cuts up a big ham and tells her family about being on top of the food chain: “Animals that get eaten are supposed to get eaten, and sometimes there were just too many of them anyway.” It’s a 1950s-era relic, though it’s also terribly sinister. Cut to those stairs in the woods, where Smart Mouth (Linden Proco) comes hopping down followed by somebody else; someone wearing a suit. Certainly not the creature that Zoe (Holland Roden) saw recently. Speaking of Zoe, she’s gone, and Alice (Olivia Luccardi) is trying to call her but getting no answer. The younger sister thinks about a year prior. She woke to find her older sister standing on her bed. Zoe says only “father time,” and soon a nasty bug comes out from under the bed. She traps it in a glass then holds it out at Alice, taunting her. They’ve got bigger problems than bugs, though. Mom’s deteriorating fast. Back in current day, Louise (Krisha Fairchild) asks Alice to show her where she met Joseph Peach (Rutger Hauer). So, they’re headed down to Butcher’s Block. Alice mentions about being there with Zoe, who saw something else in the woods. The two women go out into the park, where the playground stands overgrown. Louise tells the younger woman about 1952, when cops came knocking on the Peachs door; nobody was there. Moreover, we find out that Peach’s daughters were killed in Butcher’s Block. After that people would go missing around there, too. Later on people “found something in the basement” – nobody knows what – and the mansion was set ablaze. Alice thinks she sees Izzy running through the grass. However, it’s actually little Smart Mouth. The tiny man soon joins up with the man in the suit, and they hold hands. Soon enough, they walk up those stairs disappearing behind the door. Alice sees the staircase, it draws her near, like she’s being pulled towards the steps. A piercing noise knocks her to the ground. Then the staircase vanishes. The one who wakes Alice is none other than old Mr. Peach. The man wants to have a chat about theology, including some maggots. Joseph’s curious whether Alice believes in God, if she worships him, so on. He goes on to mention Zoe and claims she’ll be fine. And he also vanishes. We see Alice call the cops and tell them about the staircase. Naturally, they don’t believe much of what she’s saying. Although it’s good to have Louise there, she criticises the “tone of voice” of the cops when they treat her young tenant poorly. When Louise and Alice get back to the house they stumble onto Zoe – the older Woods sister claims to be “cured.” Zoe tells Alice about meeting a man after she left. She talks of an experience while being there with him. Ultimately, she was fixed through a vision. Yet the younger sister knows there are sinister forces around Butcher’s Block, and now she can’t convince the older one of that. As it turns out, there’s a “side effect” to the Peach cure, too. All the while there are strange things happening all around Butcher’s Block. That eerie creature from the staircase in “Insidious Onset” has little Izzy, and it’s preparing her in a pentagram-like circle: looks like maybe he’s making a mask of her face. The scissor lady (Paula Boudreau) is out on the streets, doing whatever it is she does, when suddenly she sees a man in a suit coming after her. The man runs her right into a police car and uses a sort of stun gun on her. The cops arrest him, though he doesn’t seem too worried – his name’s Robert Peach (Andreas Apergis); yeah, one of the famous family. And lord, is Bob a creep. Back in his cell, Mr. Peach not only murders the man in there with him, he’s also pulling organs right out of the dude’s stomach and chewing on them. Really upsets poor Officer Luke Vanczyk (Brandon Scott) when he walks in. His father – Chief Vanczyk (Tyrone Benskin) – sees him as too “sensitive,” urging him to take a couple days off. Well Luke finds out that the cannibalistic Peach is being brought back to Medallion Park. There, Bob’s given back his suit, and he wanders through the woods back towards the playground. Luke follows him further until they reach the stairs, where Joseph awaits. The two Peach men walk up the staircase towards the door together. At home, Zoe’s having trouble. She sees herself picking up Louise’s hairless cat and biting into it like food. Downstairs, Louise and Alice are going over the disappearances of Butcher’s Block and what the former believes about the neighbourhood. The graffiti murals around with the “white face” and “bloody mouth” are depictions of the Butcher, who many claim is Peach after he went mad over the murders of his daughters. That’s when the two women discover the front door open and a letter left on the table addressed to the Woods sisters – it’s from the Peach family, an invitation for dinner in the park. That night, Alice and Louise go out to the park where the Peach family sit around a table lit up with candles. Suddenly, they’re inside the old mansion. Joseph explains he did “a procedure” on Zoe. The Peachs take revenge on the residents of Butcher’s Block for the murder of the two sisters; they eat them, yum, yum. “Bless this family, and bless the bones of our enemies.” Eventually, Joseph shows Alice the procedure he did on her sister. He paralyses her, then uses a crude contraption to open up the top of her skull. NO BIG DEAL, RIGHT?! Zoe was pretty much right when she said it was like ole Joey extracted part of her. He literally prods at her brain with his fingers, searching for her “deepest trauma” involving mom’s horrible breakdown. Turns out, mom stabbed Zoe in the stomach on that fateful day. Joe finds “the beast” in her brain, the schizophrenia hiding and waiting to come out. Once it’s over, Alice is back in the world again, in the woods at the park. She’s sitting at a table full of rancid human meat, flies buzzing, to her and Louise’s horror. They’ve been sitting at the table of a family of cannibals all night.

JESUS, this episode was wild. Really took things to an entirely new level of weird, which is something I expect after Season 1 and 2 of Channel Zero. Cannot get enough! “All You Ghost Mice” is next week. Channel Zero – Butcher’s Block, Episode 2: “Father Time” Syfy's Channel Zero Season 3, Episode 2: "Father Time" Directed by Arkasha Stevenson…

[flauntpage]

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks

The first couple weeks of the 2017-18 NBA season have been more fun, unpredictable, and mind-boggling than anyone could've guessed. After Gordon Hayward's injury, the Boston Celtics look like they'll never lose again, Aaron Gordon appears to be a budding All-Star, and Cleveland Cavaliers general manager Koby Altman should probably consider blowing everything up and starting all over (kidding!).

Seriously, though, the season is quickly shaping into an entertaining adventure nobody saw coming: an ongoing drama between belief and skepticism. Between the Little Engine That Could and Small Sample Size Mountain. Let's take a closer look.

1. Chandler Parsons Looks (Relatively) Phenomenal

Heading into this season, expectations surrounding Chandler Parsons—at 29, post-several significant knee surgeries, after a year in which no player in the entire league (except, um, his own teammate Andrew Harrison) shot the ball worse—were lower than they will be for Netflix's inevitable rollout of Stranger Things 7. But instead of hobbling around as a $23 million ball mover, Parsons is one of the most efficient players in the entire league—last night's 0-for-4 outing against the Orlando Magic notwithstanding—and possesses its lowest defensive rating.

(When he's on the court, Memphis performs like a 79-win team! When he's off, only 29.)

Parsons isn't blowing by defenders (unless they're named "Frank Kaminsky"), but has finally rediscovered some confidence in his shot after starting the season with a petrified look on his face every time someone passed him the ball. He's averaging more points per 36 minutes than ever, and has spent nearly all his time at the four (a smart, new development that's partly due to JaMychal Green's ankle injury).

Parsons recorded two dunks in his first 100 minutes after a grand total of three in 674 minutes last season. On one play against the Charlotte Hornets, he grabbed a defensive rebound, leisurely dribbled to the top of the arc, and launched an open three. It clanged off the front iron, but that's still an encouraging level of comfort to see from a guy who was booed by his own fanbase a couple weeks ago.

What does all this mean for the Memphis Grizzlies? Parsons has only logged 19 minutes beside Mike Conley and Marc Gasol, and in that time they were outscored by 14.2 points per 100 possessions. But if they can gel some on an upcoming five-game road trip, and Parsons is able to sustain some of his efficiency in a larger role without suffering any health-related setbacks, there's a very good chance this team can not only qualify for the playoffs, but make some genuine noise once they're there.

2. Big Men and Closeouts

This might seem obvious, but with even more traditional centers stepping behind the three-point line this year, the guys asked to stop them are also drifting towards the perimeter more than they used to. The following qualifies as anecdotal evidence within a small sample size, but according to NBA.com Dwight Howard is contesting 2.7 threes per game this season, up from 1.5 last year. Marc Gasol is at 3.7 three-point contests, and last year he averaged 1.8. Steven Adams contested 2.7 threes last year and now he's at 3.6.

Again, these numbers are circumstantial—reliant on minutes, opponents, and scheme in a tiny sample size—and should be read with a grain of salt. Some centers (like Rudy Gobert and DeMarcus Cousins) haven't seen any uptick at all. But what matters here is the reminder that as NBA offenses continue to evolve, individual defenders are being forced to need to sharpen tools they barely used to need.

Centers who bite at Joel Embiid's pump fake, or wildly race out at Brook Lopez with no plan other than to run him off the line, put pressure on help defenders who're forced to either foul, take a very painful charge, or desert their own assignment and surrender an open look elsewhere.

Sprinting to a dead stop and then trying to laterally stick with a ball-handler is incredibly difficult, but in today's NBA this is what once-plodding seven-footers have to do if they want to stay on the floor.

3. Apologies to Jakob Poeltl

I don't think my opinion on a prospect has ever shifted faster than it has with Jakob Poeltl. It was unclear watching him last year how a seven-footer who can't shoot and doesn't possess leap-off-the-screen athleticism could carve out a meaningful role on a winning team.

This opinion was bad. Poeltl is awesome. Not only is he a putback monster who can control the offensive glass against the right matchup (Toronto's offensive rebound rate is 9.3 percent higher with him in the game), but the 22-year-old has also proven to be an agile pick-and-roll finisher, with touch and strength around the rim. His defense is phenomenal, too, particularly when switching out on the perimeter. Poeltl keeps one hand high to bother the shooter's vision, swivels his hips, and slides step for step.

This is valuable, but thanks to Jonas Valanciunas and Bebe Nogueira, Poeltl's playing time isn't as high as his skills suggest it should be.

4. Philly's Expanding Playbook

It's oh so very early, but according to Synergy Sports, the Philadelphia 76ers boast the NBA's most efficient offense after a timeout. This is a massive leap from last year, when, well, they came in dead last, averaging a measly 0.819 points per possession. Some of this is thanks to Brett Brown's willingness to experiment with the most talented and complementary roster he's ever had, and some is just because said talent is able to savage defenses that aren't as focused as they should be.

Ben Simmons is as perceptive as he is physically imposing; the 21-year-old has already figured out how to make opponents pay when they don't execute as tightly as they should (or when they're simply unable to squeeze the ball out of his hands).

After an Iverson cut towards the left wing, Simmons attacks away from the screen once he notices that Dallas Mavericks big Dwight Powell is hugged up on Amir Johnson instead of in position to ice the pick-and-roll.

The next play starts the same, with Simmons once again opening things up by cutting across the elbow. But instead of Johnson setting a screen, Joel Embiid posts up on the left block while three other Sixers (who're all respectable outside threats) clear out to the weakside. Trevor Ariza isn't in position to force Simmons towards the sideline, so the phenom behaves like a phenom and instead plows into the middle towards an open lane.

These two positive results come off action that isn't especially creative. But Brown is smart enough to realize that sometimes all he has to do is get out of the way. Wind up your franchise player, point him towards a simple two-man action, then let him wreak some havoc. Simmons's ability to read and react at warp speed is one of the many unteachable gifts he already has, and the scheme that can slow him down might not currently exist.

5. Is Ricky Rubio Finally Evolving?

Watch what happens when a defense goes out of its way to prevent Rubio from shooting the ball.

As he spins middle off Gobert's screen, Brandon Ingram leaves Joe Ingles (you know, the guy who made 44.1 percent of his threes last season and is even more accurate this year) to stunt and force a pass. The ball is eventually swung to the opposite corner, where Rodney Hood drills an open look.

This is probably more due to an antsy 20-year-old trying to make a play than a tactical decision from Lakers head coach Luke Walton, but it hints at a reality many thought we'd never see. Rubio is making shots. What's even more impressive than him making 38 percent of his threes (and a completely unsustainable 54 percent of his long twos) is a newfound bravery attached to his shot selection.

Rubio's three-point rate is currently 16.7 percent higher than his career average. Above-the-break treys are still all over the place and he still can't finish at the rim, but a willingness to fire away could change how defenses treat him over the course of the season. Off reputation alone, Rubio's gravity won't ever sniff most of his contemporaries, but an ability to make defenders pay every now and again is significant.

(Also, he has the best hair in the league.)

On Wednesday, Rubio finished with 30 points (three short of his career high) on 17 shots. For just one moment, imagine an alternate reality where these developments are taking place on a Jazz roster that also has a healthy Gordon Hayward and Derrick Favors nearly back to the borderline-All-Star plane where he ascended before injuries weakened his antithetical impact. Is that the second or third-best team in the Western Conference? Does a Rubio, Hood, Hayward, Favors, Gobert lineup make the Warriors sweat?

6. Reminder: Giannis is Huge!

The sight never gets old. In the opening few minutes of Milwaukee's blowout loss against Oklahoma City on Tuesday night, Giannis glided around the floor as a taller, stronger, longer, version of all the various wing defenders employed by the Thunder. It was funny, watching OKC's fundamental identity and nightly advantage look so delicate standing beside the NBA's very own Cloverfield. On the same court as Giannis, Paul George, Andre Roberson, and Jerami Grant looked like raptors flailing around in Jurassic Park's final scene.

7. Reminder: De'Aaron Fox is Fast!

Keep an eye on the shot clock.

8. Is 2017-18 Kemba Walker About to Become 2016-17 Isaiah Thomas?

Meaning, are we in store for a second unexpected leap from a spunky Eastern Conference point guard, one season after it felt like they already spilled out all they had to offer? Earlier this week, Walker ranked third in fourth-quarter scoring (he's now at 10th, with a number that would be top five last season), has never been more efficient from inside or outside the arc, and has damn near doubled his free-throw rate.

Walker has been fantastic inside the paint, and the Charlotte Hornets look deprived of all five senses when he's off the court. This is somewhat due to the fact that they don't currently have a backup point guard, but Charlotte is still an unbelievable 33.6 points per 100 possessions better when he's in there.

There's a jumpy, unpredictability to Walker's game right now. On one recent possession against the Memphis Grizzlies, Walker pushed the ball in transition and nearly penetrated beneath the basket before he decided to pump the brakes and dribble back out to set up the offense. But once he realized no Grizzlies were nearby to escort him to the perimeter, he curled baseline and knocked down a wide open jumper. Splash.

With more pressure to shoulder a heavier load after Nicolas Batum went down in the preseason, Walker is playing with an unseen self-belief that's steadily elevating his game even higher than last year's All-Star campaign showed it could go. Taming a tiger is less complicated than corralling him off a high screen right now. He's a virtual lock to make his second-straight All-Star team.

9. The New York Knicks are Rebounding the Shit out of the Ball

Photo by Wendell Cruz - USA TODAY Sports

Remember when the Knicks were mocked for constructing a roster that essentially barred Kristaps Porzingis from spending any time at center (only three percent of his minutes have been at that position this year, down from 21 percent last year)? Well, even after three-straight wins against the Brooklyn Nets, Cleveland Cavaliers, and Denver Nuggets that took place before they were slapped back to Earth by the Houston Rockets, these personnel decisions probably still weren't the way to go.

But what those personnel choices have done is help New York formulate a fun, possibly sustainable (?) Porzingis + Putbacks identity. With Carmelo Anthony out of the picture, Porzingis has spent the opening chapter of his third season mushrooming into an unguardable beanstock. Only Giannis, Boogie, and Steph Curry are averaging more points than Porzingis. Zero players have a higher usage rate.

Instead of spacing issues caused by the likes of Enes Kanter and Kyle O'Quinn, those two have butchered teams on the glass. The Knicks rank second in offensive rebound rate and third in total rebound rate. While almost every other team around the league is downsizing, New York has firmly positioned their 7'3" franchise player at the four. And, relative to some depressing expectations, it's working!

10. I Can't Wait for the Atlanta Hawks to be Good

If you've happened to catch any recent Hawks game at Philips Arena, you might remember sideline reporter Andre Aldridge posted up at a brand new bar that just opened along the court's baseline. It looks like the most amazing place on Earth.

The team is horrible, but have openly cuddled up beside a full-on rebuild that should (if all goes according to plan) make Philips Arena one of the NBA's most lively atmospheres a few years down the road. Until then, Dominique Wilkins and Bob Rathbun need to broadcast every home game games directly from the bar.

11. Let's Trade Jamal Murray for Kyle Lowry

The likeliness of a trade involving these two players is microscopic—the idea disintegrates if the Toronto Raptors and Denver Nuggets both look like solid playoff teams in late January (Lowry can't be dealt until that month)—so I won't spend too much time rationalizing why I think it should happen.

But it sorta makes sense! Big picture, Toronto has a rapidly progressing core simmering beneath its veteran, All-Star-caliber contributors. The aforementioned Poeltl, rookie OG Anunoby, recently signed Norm Powell, and intriguing rotation players like Delon Wright and Pascal Siakam have the future looking solid.

They're successfully rebuilding on the fly while Lowry, DeMar DeRozan, and Serge Ibaka begin to decline on big-money contracts. Trading (at least) one of those three for valuable assets would punt meaningful playoff contention from 2018-2020, but allow continuity to accelerate within a new, modernized offensive system.

If they can somehow land someone with Murray's upside and turn him into their new franchise player, the Raptors would seamlessly glide from a stagnant also-ran to a promising up-and-comer. Dwane Casey has already relented a bit, playing lineups that feature four or five young pups at the same time.

The main holdup here, besides contractual issues that make matching money a little difficult with these two teams, is Denver's cooperation. Why the hell would they give up on a 20-year-old who defends his position and may own the most invaluable offensive trait in basketball: an ability to knock down pull-up threes at a reliable rate?

Denver is almost an inverse of the Raptors. Both teams are operating on two timelines, but the Nuggets are more clearly loaded to do damage five years from now. Nikola Jokic is 22, Gary Harris just turned 23, and Emmanuel Mudiay (who's made 45.5 percent of his threes this year!) is 21. Common sense says "wait." But Paul Millsap's decision to climb aboard turns maximizing the present into a conversation.

Lowry has been pretty bad this year, but he's still one of the five or six most effective all-around players at his position. Imagine how he'd look next to Jokic and Millsap. How much better would Denver be if he's there this season and next?

Again, a trade like this is extremely complicated and would dramatically shift the direction of two franchises that seem to be content with where they are. But the word impossible doesn't exist in today's NBA.

12. Can Rashad Vaughn Maybe Become a Thing?

Vaughn (who recently said "that's what we lived for" in reference to the McGriddle sandwich) entered his third season with one foot in the league and the other on a banana peel. He logged a grand total of four and a half minutes in Milwaukee's first four games (during which he was trade bait) before draining four threes in an 11-point win against the Hawks.

On Halloween, the team decided not to pick up his fourth-year option, making Vaughn an unrestricted free agent this summer. For a team that has little financial flexibility going forward, completely whiffing on a first-round pick can have painful consequences. Giannis is clearly ready to win now, and the Eastern Conference is begging someone to usurp the Cavaliers.

As Malcolm Brogdon, Jabari Parker, and Khris Middleton each become eligible for a significant pay raise in the next couple summers, the pressure will be on Milwaukee's front office to complement their franchise megastar with a championship-caliber supporting cast before he can flee as a free agent.

On paper, Vaughn is an ideal puzzle piece: a 6'6" three-point threat who may one day be able to reliably knock down threes, make plays when the ball is swung his way, and threaten defenses by pulling up off a dribble hand-off or initiating his own pick-and-roll. Maybe the Bucks believe waiting to see if Vaughn pans out is a waste of everybody's time, especially now that Tony Snell already fills the role he was meant for.

But money issues constrict ways in which Milwaukee can improve from the outside. Internal improvement is key. Vaughn's team option feels negligible now, but giving up on him so soon may come back to haunt this team in one way or another.

13. The Spurs are Perfect Even When They're Not

Even though Patty Mills' game-tying three didn't fall, San Antonio's execution of this elevators action at the end of a recent loss against the Indiana Pacers exemplified why they're the coolest cucumbers around.

Everything about this is ideal...until the ball leaves his fingertips.

14. Jordan Clarkson's Usage Rate is Higher than Anthony Davis, Russell Westbrook, and Just About Everybody Else

To suggest Clarkson has made the most of his reduced playing time is to suggest that Kendrick Lamar sometimes steals the spotlight when he's on other people's songs. In ten fewer minutes than he averaged last year, Clarkson is averaging the same amount of points, knocking down threes at a more accurate clip, posting the highest assist rate of his career, and, generally standing out as a quality contributor off Los Angeles' bench. (He launched six threes in 14 freaking minutes against the Toronto Raptors!)

He's efficient for the very first time despite his usage percentage soaring into the rarified air normally reserved for All-Stars. Some of this is because he's the only shot creator on the floor, often paired with the likes of Corey Brewer, Kyle Kuzma, and Josh Hart. And some of it's because he's been instructed to attack. It's too early to speculate whether this is a breakout campaign or just an early-season surge, but Clarkson's production is flying under the radar in a city that thinks Lonzo Ball is the only player who ever lived.

15. Al Horford is Playing Better Defense Than Everybody Else

The Boston Celtics have the best defense in the NBA because Al Horford is playing like its best defender. When he's off the floor they guard like a bottom-10 unit, but when he's out there, nailed down as a human lighthouse guiding Boston's young pack of swarming athletes everywhere they should go, the Celtics are well-choreographed misery.

Individually, the overwhelming talent Horford has had to corral is beyond impressive: Giannis (twice), Ben Simmons, LaMarcus Aldridge, Kristaps Porzingis, and Kevin Love. All opponents are shooting just 56.2 percent at the rim when Horford is on the floor. When he rests, that number spikes all the way up to 74.6 percent. The difference ranks in the 98th percentile among players at his position, according to Cleaning the Glass).

For the fleeting minority that still scoffs at Horford's occasional humdrum box score, and are fed up with the Ambien-akin side-effects commonly linked to what happens after repeated exposure to negated entry passes, crisp high screens, and perfect execution of myriad pick-and-roll coverages, Horford remains an overpaid waste. Nearly a dozen years of evidence proves they're wrong, and this year he's definitively worth every penny.

Using priceless instincts, flawless habits, and a wingspan that allows him to cover more ground than anyone his size should (only seven players contest more shots every game, per NBA.com), Horford has glued himself inside the all-too-early Defensive Player of the Year (pseudo-MVP?) conversation. He shouldn't leave it anytime soon.

16. Tristan Thompson is a Black Eye on Cleveland's Bloody Face

The Cavaliers have dropped five of their last six games, with all five losses coming up against teams few, if anybody, projected to make the playoffs. Life is rough. But on a team with defensive woes that are as much due to mental indifference as they are physical fragility, Thompson's struggles across the board are particularly worrisome.

Two years ago, the Cavaliers allowed 101.7 points per 100 possessions with Thompson on the floor. This season, his defensive rating is 111.2. His minutes are down, his confidence is low, and his offensive role is non-existent. It's obviously possible for the Cavaliers to bounce back after Isaiah Thomas returns and LeBron James starts to feel like a superhero.

But up until he suffered a calf injury against the Indiana Pacers that will sideline him about a month, Thompson was a non-threat off the ball who launched more long twos than he ever should. If James leaves in free agency this summer, the $36 million Thompson is owed over the next two years turn that contract into one of the league's roughest (from Cleveland's perspective!) agreements.

To be fair, once he's healthy, Thompson's numbers should stabilize once Cleveland works an actual point guard into their rotation. Teams were able to switch James-Thompson pick-and-rolls, and the sliver of opportunity born from that action mainly arrived after a mistake. Here's an example, as miscommunication between Jrue Holiday and Dante Cunningham leads to an easy dunk.

17. Dillon Brooks is Found Money

I wonder how a lucky a front office feels whenever they draft someone 45th overall and then immediately watch him flourish in consequential ways. Is this like finding a $20 bill in your back pocket or hearing your train approach the second you descend onto a subway platform?

The Memphis Grizzlies have had their fair share of first-round blunders, but scoring with guys like Brooks has helped keep this organization afloat, stiff-arming a rebuild further out than it probably should be.

I don't have much to say about Brooks. He seems to be a cagey one-on-one defender, someone who's relentless and difficult to screen. That's nice. He's also committed a bunch of rookie mistakes and isn't really making his threes. But the fact that he's averaging 30 minutes per game on one of the league's most pleasant surprises is telling.

The value of a second-round pick is never more clear than in transcendent figures like Manu Ginobili or Draymond Green, but they still feel like an undervalued commodity. Think about how different the Los Angeles Clippers might look today if they drafted someone like Brooks a few years ago?

Plucking a helpful contributor in the second round takes quite a bit of luck, but some teams have an ability to carve their own more often than others.

18. Your Weekly Reminder that the Golden State Warriors are Unfair

Kevin Durant is shooting 49 percent from behind the three-point line, and his three-point rate has never been higher.

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks published first on http://ift.tt/2pLTmlv

[amtushinfosolutionspage]

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks

The first couple weeks of the 2017-18 NBA season have been more fun, unpredictable, and mind-boggling than anyone could’ve guessed. After Gordon Hayward’s injury, the Boston Celtics look like they’ll never lose again, Aaron Gordon appears to be a budding All-Star, and Cleveland Cavaliers general manager Koby Altman should probably consider blowing everything up and starting all over (kidding!).

Seriously, though, the season is quickly shaping into an entertaining adventure nobody saw coming: an ongoing drama between belief and skepticism. Between the Little Engine That Could and Small Sample Size Mountain. Let’s take a closer look.

1. Chandler Parsons Looks (Relatively) Phenomenal

Heading into this season, expectations surrounding Chandler Parsons—at 29, post-several significant knee surgeries, after a year in which no player in the entire league (except, um, his own teammate Andrew Harrison) shot the ball worse—were lower than they will be for Netflix’s inevitable rollout of Stranger Things 7. But instead of hobbling around as a $23 million ball mover, Parsons is one of the most efficient players in the entire league—last night’s 0-for-4 outing against the Orlando Magic notwithstanding—and possesses its lowest defensive rating.

(When he’s on the court, Memphis performs like a 79-win team! When he’s off, only 29.)

Parsons isn’t blowing by defenders (unless they’re named “Frank Kaminsky”), but has finally rediscovered some confidence in his shot after starting the season with a petrified look on his face every time someone passed him the ball. He’s averaging more points per 36 minutes than ever, and has spent nearly all his time at the four (a smart, new development that’s partly due to JaMychal Green’s ankle injury).

Parsons recorded two dunks in his first 100 minutes after a grand total of three in 674 minutes last season. On one play against the Charlotte Hornets, he grabbed a defensive rebound, leisurely dribbled to the top of the arc, and launched an open three. It clanged off the front iron, but that’s still an encouraging level of comfort to see from a guy who was booed by his own fanbase a couple weeks ago.

What does all this mean for the Memphis Grizzlies? Parsons has only logged 19 minutes beside Mike Conley and Marc Gasol, and in that time they were outscored by 14.2 points per 100 possessions. But if they can gel some on an upcoming five-game road trip, and Parsons is able to sustain some of his efficiency in a larger role without suffering any health-related setbacks, there’s a very good chance this team can not only qualify for the playoffs, but make some genuine noise once they’re there.

2. Big Men and Closeouts

This might seem obvious, but with even more traditional centers stepping behind the three-point line this year, the guys asked to stop them are also drifting towards the perimeter more than they used to. The following qualifies as anecdotal evidence within a small sample size, but according to NBA.com Dwight Howard is contesting 2.7 threes per game this season, up from 1.5 last year. Marc Gasol is at 3.7 three-point contests, and last year he averaged 1.8. Steven Adams contested 2.7 threes last year and now he’s at 3.6.

Again, these numbers are circumstantial—reliant on minutes, opponents, and scheme in a tiny sample size—and should be read with a grain of salt. Some centers (like Rudy Gobert and DeMarcus Cousins) haven’t seen any uptick at all. But what matters here is the reminder that as NBA offenses continue to evolve, individual defenders are being forced to need to sharpen tools they barely used to need.

Centers who bite at Joel Embiid’s pump fake, or wildly race out at Brook Lopez with no plan other than to run him off the line, put pressure on help defenders who’re forced to either foul, take a very painful charge, or desert their own assignment and surrender an open look elsewhere.

Sprinting to a dead stop and then trying to laterally stick with a ball-handler is incredibly difficult, but in today’s NBA this is what once-plodding seven-footers have to do if they want to stay on the floor.

3. Apologies to Jakob Poeltl

I don’t think my opinion on a prospect has ever shifted faster than it has with Jakob Poeltl. It was unclear watching him last year how a seven-footer who can’t shoot and doesn’t possess leap-off-the-screen athleticism could carve out a meaningful role on a winning team.

This opinion was bad. Poeltl is awesome. Not only is he a putback monster who can control the offensive glass against the right matchup (Toronto’s offensive rebound rate is 9.3 percent higher with him in the game), but the 22-year-old has also proven to be an agile pick-and-roll finisher, with touch and strength around the rim. His defense is phenomenal, too, particularly when switching out on the perimeter. Poeltl keeps one hand high to bother the shooter’s vision, swivels his hips, and slides step for step.

This is valuable, but thanks to Jonas Valanciunas and Bebe Nogueira, Poeltl’s playing time isn’t as high as his skills suggest it should be.

4. Philly’s Expanding Playbook

It’s oh so very early, but according to Synergy Sports, the Philadelphia 76ers boast the NBA’s most efficient offense after a timeout. This is a massive leap from last year, when, well, they came in dead last, averaging a measly 0.819 points per possession. Some of this is thanks to Brett Brown’s willingness to experiment with the most talented and complementary roster he’s ever had, and some is just because said talent is able to savage defenses that aren’t as focused as they should be.

Ben Simmons is as perceptive as he is physically imposing; the 21-year-old has already figured out how to make opponents pay when they don’t execute as tightly as they should (or when they’re simply unable to squeeze the ball out of his hands).

After an Iverson cut towards the left wing, Simmons attacks away from the screen once he notices that Dallas Mavericks big Dwight Powell is hugged up on Amir Johnson instead of in position to ice the pick-and-roll.

The next play starts the same, with Simmons once again opening things up by cutting across the elbow. But instead of Johnson setting a screen, Joel Embiid posts up on the left block while three other Sixers (who’re all respectable outside threats) clear out to the weakside. Trevor Ariza isn’t in position to force Simmons towards the sideline, so the phenom behaves like a phenom and instead plows into the middle towards an open lane.

These two positive results come off action that isn’t especially creative. But Brown is smart enough to realize that sometimes all he has to do is get out of the way. Wind up your franchise player, point him towards a simple two-man action, then let him wreak some havoc. Simmons’s ability to read and react at warp speed is one of the many unteachable gifts he already has, and the scheme that can slow him down might not currently exist.

5. Is Ricky Rubio Finally Evolving?

Watch what happens when a defense goes out of its way to prevent Rubio from shooting the ball.

As he spins middle off Gobert’s screen, Brandon Ingram leaves Joe Ingles (you know, the guy who made 44.1 percent of his threes last season and is even more accurate this year) to stunt and force a pass. The ball is eventually swung to the opposite corner, where Rodney Hood drills an open look.

This is probably more due to an antsy 20-year-old trying to make a play than a tactical decision from Lakers head coach Luke Walton, but it hints at a reality many thought we’d never see. Rubio is making shots. What’s even more impressive than him making 38 percent of his threes (and a completely unsustainable 54 percent of his long twos) is a newfound bravery attached to his shot selection.

Rubio’s three-point rate is currently 16.7 percent higher than his career average. Above-the-break treys are still all over the place and he still can’t finish at the rim, but a willingness to fire away could change how defenses treat him over the course of the season. Off reputation alone, Rubio’s gravity won’t ever sniff most of his contemporaries, but an ability to make defenders pay every now and again is significant.

(Also, he has the best hair in the league.)

On Wednesday, Rubio finished with 30 points (three short of his career high) on 17 shots. For just one moment, imagine an alternate reality where these developments are taking place on a Jazz roster that also has a healthy Gordon Hayward and Derrick Favors nearly back to the borderline-All-Star plane where he ascended before injuries weakened his antithetical impact. Is that the second or third-best team in the Western Conference? Does a Rubio, Hood, Hayward, Favors, Gobert lineup make the Warriors sweat?

6. Reminder: Giannis is Huge!

The sight never gets old. In the opening few minutes of Milwaukee’s blowout loss against Oklahoma City on Tuesday night, Giannis glided around the floor as a taller, stronger, longer, version of all the various wing defenders employed by the Thunder. It was funny, watching OKC’s fundamental identity and nightly advantage look so delicate standing beside the NBA’s very own Cloverfield. On the same court as Giannis, Paul George, Andre Roberson, and Jerami Grant looked like raptors flailing around in Jurassic Park‘s final scene.

7. Reminder: De’Aaron Fox is Fast!

Keep an eye on the shot clock.

8. Is 2017-18 Kemba Walker About to Become 2016-17 Isaiah Thomas?

Meaning, are we in store for a second unexpected leap from a spunky Eastern Conference point guard, one season after it felt like they already spilled out all they had to offer? Earlier this week, Walker ranked third in fourth-quarter scoring (he’s now at 10th, with a number that would be top five last season), has never been more efficient from inside or outside the arc, and has damn near doubled his free-throw rate.

Walker has been fantastic inside the paint, and the Charlotte Hornets look deprived of all five senses when he’s off the court. This is somewhat due to the fact that they don’t currently have a backup point guard, but Charlotte is still an unbelievable 33.6 points per 100 possessions better when he’s in there.

There’s a jumpy, unpredictability to Walker’s game right now. On one recent possession against the Memphis Grizzlies, Walker pushed the ball in transition and nearly penetrated beneath the basket before he decided to pump the brakes and dribble back out to set up the offense. But once he realized no Grizzlies were nearby to escort him to the perimeter, he curled baseline and knocked down a wide open jumper. Splash.

With more pressure to shoulder a heavier load after Nicolas Batum went down in the preseason, Walker is playing with an unseen self-belief that’s steadily elevating his game even higher than last year’s All-Star campaign showed it could go. Taming a tiger is less complicated than corralling him off a high screen right now. He’s a virtual lock to make his second-straight All-Star team.

9. The New York Knicks are Rebounding the Shit out of the Ball

Photo by Wendell Cruz – USA TODAY Sports

Remember when the Knicks were mocked for constructing a roster that essentially barred Kristaps Porzingis from spending any time at center (only three percent of his minutes have been at that position this year, down from 21 percent last year)? Well, even after three-straight wins against the Brooklyn Nets, Cleveland Cavaliers, and Denver Nuggets that took place before they were slapped back to Earth by the Houston Rockets, these personnel decisions probably still weren’t the way to go.

But what those personnel choices have done is help New York formulate a fun, possibly sustainable (?) Porzingis + Putbacks identity. With Carmelo Anthony out of the picture, Porzingis has spent the opening chapter of his third season mushrooming into an unguardable beanstock. Only Giannis, Boogie, and Steph Curry are averaging more points than Porzingis. Zero players have a higher usage rate.

Instead of spacing issues caused by the likes of Enes Kanter and Kyle O’Quinn, those two have butchered teams on the glass. The Knicks rank second in offensive rebound rate and third in total rebound rate. While almost every other team around the league is downsizing, New York has firmly positioned their 7’3″ franchise player at the four. And, relative to some depressing expectations, it’s working!

10. I Can’t Wait for the Atlanta Hawks to be Good

If you’ve happened to catch any recent Hawks game at Philips Arena, you might remember sideline reporter Andre Aldridge posted up at a brand new bar that just opened along the court’s baseline. It looks like the most amazing place on Earth.

The team is horrible, but have openly cuddled up beside a full-on rebuild that should (if all goes according to plan) make Philips Arena one of the NBA’s most lively atmospheres a few years down the road. Until then, Dominique Wilkins and Bob Rathbun need to broadcast every home game games directly from the bar.

11. Let’s Trade Jamal Murray for Kyle Lowry

The likeliness of a trade involving these two players is microscopic—the idea disintegrates if the Toronto Raptors and Denver Nuggets both look like solid playoff teams in late January (Lowry can’t be dealt until that month)—so I won’t spend too much time rationalizing why I think it should happen.

But it sorta makes sense! Big picture, Toronto has a rapidly progressing core simmering beneath its veteran, All-Star-caliber contributors. The aforementioned Poeltl, rookie OG Anunoby, recently signed Norm Powell, and intriguing rotation players like Delon Wright and Pascal Siakam have the future looking solid.

They’re successfully rebuilding on the fly while Lowry, DeMar DeRozan, and Serge Ibaka begin to decline on big-money contracts. Trading (at least) one of those three for valuable assets would punt meaningful playoff contention from 2018-2020, but allow continuity to accelerate within a new, modernized offensive system.

If they can somehow land someone with Murray’s upside and turn him into their new franchise player, the Raptors would seamlessly glide from a stagnant also-ran to a promising up-and-comer. Dwane Casey has already relented a bit, playing lineups that feature four or five young pups at the same time.

The main holdup here, besides contractual issues that make matching money a little difficult with these two teams, is Denver’s cooperation. Why the hell would they give up on a 20-year-old who defends his position and may own the most invaluable offensive trait in basketball: an ability to knock down pull-up threes at a reliable rate?

Denver is almost an inverse of the Raptors. Both teams are operating on two timelines, but the Nuggets are more clearly loaded to do damage five years from now. Nikola Jokic is 22, Gary Harris just turned 23, and Emmanuel Mudiay (who’s made 45.5 percent of his threes this year!) is 21. Common sense says “wait.” But Paul Millsap’s decision to climb aboard turns maximizing the present into a conversation.

Lowry has been pretty bad this year, but he’s still one of the five or six most effective all-around players at his position. Imagine how he’d look next to Jokic and Millsap. How much better would Denver be if he’s there this season and next?

Again, a trade like this is extremely complicated and would dramatically shift the direction of two franchises that seem to be content with where they are. But the word impossible doesn’t exist in today’s NBA.

12. Can Rashad Vaughn Maybe Become a Thing?

Vaughn (who recently said “that’s what we lived for” in reference to the McGriddle sandwich) entered his third season with one foot in the league and the other on a banana peel. He logged a grand total of four and a half minutes in Milwaukee’s first four games (during which he was trade bait) before draining four threes in an 11-point win against the Hawks.

On Halloween, the team decided not to pick up his fourth-year option, making Vaughn an unrestricted free agent this summer. For a team that has little financial flexibility going forward, completely whiffing on a first-round pick can have painful consequences. Giannis is clearly ready to win now, and the Eastern Conference is begging someone to usurp the Cavaliers.

As Malcolm Brogdon, Jabari Parker, and Khris Middleton each become eligible for a significant pay raise in the next couple summers, the pressure will be on Milwaukee’s front office to complement their franchise megastar with a championship-caliber supporting cast before he can flee as a free agent.

On paper, Vaughn is an ideal puzzle piece: a 6’6″ three-point threat who may one day be able to reliably knock down threes, make plays when the ball is swung his way, and threaten defenses by pulling up off a dribble hand-off or initiating his own pick-and-roll. Maybe the Bucks believe waiting to see if Vaughn pans out is a waste of everybody’s time, especially now that Tony Snell already fills the role he was meant for.

But money issues constrict ways in which Milwaukee can improve from the outside. Internal improvement is key. Vaughn’s team option feels negligible now, but giving up on him so soon may come back to haunt this team in one way or another.

13. The Spurs are Perfect Even When They’re Not

Even though Patty Mills’ game-tying three didn’t fall, San Antonio’s execution of this elevators action at the end of a recent loss against the Indiana Pacers exemplified why they’re the coolest cucumbers around.

Everything about this is ideal…until the ball leaves his fingertips.

14. Jordan Clarkson’s Usage Rate is Higher than Anthony Davis, Russell Westbrook, and Just About Everybody Else

To suggest Clarkson has made the most of his reduced playing time is to suggest that Kendrick Lamar sometimes steals the spotlight when he’s on other people’s songs. In ten fewer minutes than he averaged last year, Clarkson is averaging the same amount of points, knocking down threes at a more accurate clip, posting the highest assist rate of his career, and, generally standing out as a quality contributor off Los Angeles’ bench. (He launched six threes in 14 freaking minutes against the Toronto Raptors!)

He’s efficient for the very first time despite his usage percentage soaring into the rarified air normally reserved for All-Stars. Some of this is because he’s the only shot creator on the floor, often paired with the likes of Corey Brewer, Kyle Kuzma, and Josh Hart. And some of it’s because he’s been instructed to attack. It’s too early to speculate whether this is a breakout campaign or just an early-season surge, but Clarkson’s production is flying under the radar in a city that thinks Lonzo Ball is the only player who ever lived.

15. Al Horford is Playing Better Defense Than Everybody Else

The Boston Celtics have the best defense in the NBA because Al Horford is playing like its best defender. When he’s off the floor they guard like a bottom-10 unit, but when he’s out there, nailed down as a human lighthouse guiding Boston’s young pack of swarming athletes everywhere they should go, the Celtics are well-choreographed misery.

Individually, the overwhelming talent Horford has had to corral is beyond impressive: Giannis (twice), Ben Simmons, LaMarcus Aldridge, Kristaps Porzingis, and Kevin Love. All opponents are shooting just 56.2 percent at the rim when Horford is on the floor. When he rests, that number spikes all the way up to 74.6 percent. The difference ranks in the 98th percentile among players at his position, according to Cleaning the Glass).

For the fleeting minority that still scoffs at Horford’s occasional humdrum box score, and are fed up with the Ambien-akin side-effects commonly linked to what happens after repeated exposure to negated entry passes, crisp high screens, and perfect execution of more pick-and-roll coverages, Horford remains an overpaid waste. Nearly a dozen years of evidence proves they’re wrong, and this year he’s definitively worth every penny.

Using priceless instincts, flawless habits, and a wingspan that allows him to cover more ground than anyone his size should (only seven players contest more shots every game, per NBA.com), Horford has glued himself inside the all-too-early Defensive Player of the Year (pseudo-MVP?) conversation. He shouldn’t leave it anytime soon.

16. Tristan Thompson is a Black Eye on Cleveland’s Bloody Face

The Cavaliers have dropped five of their last six games, with all five losses coming up against teams few, if anybody, projected to make the playoffs. Life is rough. But on a team with defensive woes that are as much due to mental indifference as they are physical fragility, Thompson’s struggles across the board are particularly worrisome.

Two years ago, the Cavaliers allowed 101.7 points per 100 possessions with Thompson on the floor. This season, his defensive rating is 111.2. His minutes are down, his confidence is low, and his offensive role is non-existent. It’s obviously possible for the Cavaliers to bounce back after Isaiah Thomas returns and LeBron James starts to feel like a superhero.

But up until he suffered a calf injury against the Indiana Pacers that will sideline him about a month, Thompson was a non-threat off the ball who launched more long twos than he ever should. If James leaves in free agency this summer, the $36 million Thompson is owed over the next two years turn that contract into one of the league’s roughest (from Cleveland’s perspective!) agreements.

To be fair, once he’s healthy, Thompson’s numbers should stabilize once Cleveland works an actual point guard into their rotation. Teams were able to switch James-Thompson pick-and-rolls, and the sliver of opportunity born from that action mainly arrived after a mistake. Here’s an example, as miscommunication between Jrue Holiday and Dante Cunningham leads to an easy dunk.

17. Dillon Brooks is Found Money

I wonder how a lucky a front office feels whenever they draft someone 45th overall and then immediately watch him flourish in consequential ways. Is this like finding a $20 bill in your back pocket or hearing your train approach the second you descend onto a subway platform?

The Memphis Grizzlies have had their fair share of first-round blunders, but scoring with guys like Brooks has helped keep this organization afloat, stiff-arming a rebuild further out than it probably should be.

I don’t have much to say about Brooks. He seems to be a cagey one-on-one defender, someone who’s relentless and difficult to screen. That’s nice. He’s also committed a bunch of rookie mistakes and isn’t really making his threes. But the fact that he’s averaging 30 minutes per game on one of the league’s most pleasant surprises is telling.

The value of a second-round pick is never more clear than in transcendent figures like Manu Ginobili or Draymond Green, but they still feel like an undervalued commodity. Think about how different the Los Angeles Clippers might look today if they drafted someone like Brooks a few years ago?

Plucking a helpful contributor in the second round takes quite a bit of luck, but some teams have an ability to carve their own more often than others.

18. Your Weekly Reminder that the Golden State Warriors are Unfair

Kevin Durant is shooting 49 percent from behind the three-point line, and his three-point rate has never been higher.

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks syndicated from http://ift.tt/2ug2Ns6

[go-ai-jay-kim-blog]

Image Completion with Deep Learning in TensorFlow

( - cited from Brandon Amos:http://bamos.github.io )

Introduction

Content-aware fill is a powerful tool designers and photographers use to fill in unwanted or missing parts of images. Image completion and inpainting are closely related technologies used to fill in missing or corrupted parts of images. There are many ways to do content-aware fill, image completion, and inpainting. In this blog post, I present Raymond Yeh and Chen Chen et al.’s paper “Semantic Image Inpainting with Perceptual and Contextual Losses,” which was just posted on arXiv on July 26, 2016. This paper shows how to use deep learning for image completion with a DCGAN. This blog post is meant for a general technical audience with some deeper portions for people with a machine learning background. I’ve added [ML-Heavy] tags to sections to indicate that the section can be skipped if you don’t want too many details. We will only look at the constrained case of completing missing pixels from images of faces. I have released all of the TensorFlow source code behind this post on GitHub at bamos/dcgan-completion.tensorflow.

We’ll approach image completion in three steps.

We’ll first interpret images as being samples from a probability distribution.

This interpretation lets us learn how to generate fake images.

Then we’ll find the best fake image for completion.

Photoshop example of automatically filling in missing image parts. (Image CC licensed, source.)

Photoshop example of automatically removing unwanted image parts. (Image CC licensed, source.)

Completions generated by what we’ll cover in this blog post. The centers of these images are being automatically generated. The source code to create this is available here. These are not curated! I selected a random subset of images from the LFW dataset.

Step 1: Interpreting images as samples from a probability distribution

How would you fill in the missing information?

In the examples above, imagine you’re building a system to fill in the missing pieces. How would you do it? How do you think the human brain does it? What kind of information would you use?

In this post we will focus on two types of information:

Contextual information: You can infer what missing pixels are based on information provided by surrounding pixels.

Perceptual information: You interpret the filled in portions as being “normal,” like from what you’ve seen in real life or from other pictures.

Both of these are important. Without contextual information, how do you know what to fill in? Without perceptual information, there are many valid completions for a context. Something that looks “normal” to a machine learning system might not look normal to humans.

It would be nice to have an exact, intuitive algorithm that captures both of these properties that says step-by-step how to complete an image. Creating such an algorithm may be possible for specific cases, but in general, nobody knows how. Today’s best approaches use statistics and machine learning to learn an approximate technique.

But where does statistics fit in? These are images.

To motivate the problem, let’s start by looking at a probability distribution that is well-understood and can be represented concisely in closed form: a normal distribution. Here’s the probability density function (PDF) for a normal distribution. You can interpret the PDF as going over the input space horizontally with the vertical axis showing the probability that some value occurs. (If you’re interested, the code to create these plots is available at bamos/dcgan-completion.tensorflow:simple-distributions.py.)

PDF for a normal distribution.

Let’s sample from the distribution to get some data. Make sure you understand the connection between the PDF and the samples.

Samples from a normal distribution.

This is a 1D probability distribution because the input only goes along a single dimension. We can do the same thing in two dimensions.

PDF and samples from a 2D normal distribution. The PDF is shown as a contour plot and the samples are overlaid.

The key relationship between images and statistics is that we can interpret images as samples from a high-dimensional probability distribution. The probability distribution goes over the pixels of images. Imagine you’re taking a picture with your camera. This picture will have some finite number of pixels. When you take an image with your camera, you are sampling from this complex probability distribution. This distribution is what we’ll use to define what makes an image normal or not. With images, unlike with the normal distributions, we don’t know the true probability distribution and we can only collect samples.

In this post, we’ll use color images represented by the RGB color model. Our images will be 64 pixels wide and 64 pixels high, so our probability distribution has

64⋅64⋅3≈12k

64

⋅

64

⋅

3

≈

12

k

dimensions.

So how can we complete images?

Let’s first consider the multivariate normal distribution from before for intuition. Given

x=1

x=

1

, what is the most probable

y

y

value? We can find this by maximizing the value of the PDF over all possible

y

y

values with

x=1

x=

1

fixed.

Finding the most probable

y

y

value given some fixed

x

x

in a multivariate normal distribution.

This concept naturally extends to our image probability distribution when we know some values and want to complete the missing values. Just pose it as a maximization problem where we search over all of the possible missing values. The completion will be the most probable image.

Visually looking at the samples from the normal distribution, it seems reasonable that we could find the PDF given only samples. Just pick your favorite statistical model and fit it to the data.

However we don’t use this method in practice. While the PDF is easy to recover for simple distributions, it’s difficult and often intractable for more complex distributions over images. The complexity partly comes from intricate conditional dependencies: the value of one pixel depends on the values of other pixels in the image. Also, maximizing over a general PDF is an extremely difficult and often intractable non-convex optimization problem.

Step 2: Quickly generating fake images

Learning to generate new samples from an unknown probability distribution

Instead of learning how to compute the PDF, another well-studied idea in statistics is to learn how to generate new (random) samples with a generative model. Generative models can often be difficult to train or intractable, but lately the deep learning community has made some amazing progress in this space. Yann LeCun gives a great introduction to one way of training generative models (adversarial training) in this Quora post, describing the idea as the most interesting idea in the last 10 years in machine learning:

Yann LeCun’s introduction to adversarial training from this Quora post.

Street Fighter analogy for adversarial networks from the EyeScream post. The networks fight each other and improve together, like two humans playing against each other in a game. Image source.

There are other ways to train generative models with deep learning, like Variational Autoencoders (VAEs). In this post we’ll only focus on Generative Adversarial Nets (GANs).

[ML-Heavy] Generative Adversarial Net (GAN) building blocks

These ideas started with Ian Goodfellow et al.’s landmark paper “Generative Adversarial Nets” (GANs), published at the Neural Information Processing Systems (NIPS) conference in 2014. The idea is that we define a simple, well-known distribution and represent it as

p

z

pz

. For the rest of this post, we’ll use

p

z

pz

as a uniform distribution between -1 and 1 (inclusively). We represent sampling a number from this distribution as

z∼

p

z

z∼pz

. If

p

z

pz

is 5-dimensional, we can sample it with one line of Python with numpy:

z = np.random.uniform(-1, 1, 5)

array([ 0.77356483,  0.95258473, -0.18345086,  0.69224724, -0.34718733])

Now this we have a simple distribution we can easily sample from, we’d like to define a function

G(z)

G(z)

that produces samples from our original probability distribution.

def G(z):

  ...

  return imageSample

z = np.random.uniform(-1, 1, 5)

imageSample = G(z)

So how do we define

G(z)

G(z)

so that it takes a vector on input and returns an image? We’ll use a deep neural network. There are many great introductions to deep neural network basics, so I won’t cover them here. Some great references that I recommend are Stanford’s CS231n course, Ian Goodfellow et al.’s Deep Learning Book, Image Kernels Explained Visually, and convolution arithmetic guide.

There are many ways we can structure

G(z)

G(z)

with deep learning. The original GAN paper proposed the idea, a training procedure, and preliminary experimental results. The idea has been greatly built on and improved. One of the most recent ideas was presented in the paper “Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks” by Alec Radford, Luke Metz, and Soumith Chintala at the International Conference on Learning Representations (ICLR, pronounced “eye-clear”) in 2016. This paper presents deep convolutional GANs (called DCGANs) that use fractionally-strided convolutions to upsample images.

What is a fractionally-strided convolution and how do they upsample images? Vincent Dumoulin and Francesco Visin’s paper “A guide to convolution arithmetic for deep learning” and conv_arithmetic project is a very well-written introduction to convolution arithmetic in deep learning. The visualizations are amazing and give great intuition into how fractionally-strided convolutions work. First, make sure you understand how a normal convolution slides a kernel over a (blue) input space to produce the (green) output space. Here, the output is smaller than the input. (If you don’t, go through the CS231n CNN section or the convolution arithmetic guide.)

Illustration of a convolution from the input (blue) to output (green). This image is from vdumoulin/conv_arithmetic.

Next, suppose that you have a 3x3 input. Our goal is to upsample so that the output is larger. You can interpret a fractionally-strided convolution as expanding the pixels so that there are zeros in-between the pixels. Then the convolution over this expanded space will result in a larger output space. Here, it’s 5x5.

Illustration of a fractionally-strided convolution from the input (blue) to output (green). This image is from vdumoulin/conv_arithmetic.

As a side-note, there are many names for convolutional layers that upsample: full convolution, in-network upsampling, fractionally-strided convolution, backwards convolution, deconvolution, upconvolution, or transposed convolution. Using the term ‘deconvolution’ for this is strongly discouraged because it’s an over-loaded term: the mathematical operation or other uses in computer vision have a completely different meaning.

Now that we have fractionally-strided convolutions as building blocks, we can finally represent

G(z)

G(z)

so that it takes a vector

z∼

p

z

z∼pz

on input and outputs a 64x64x3 RGB image.

One way to structure the generator

G(z)

G(z)

with a DCGAN. This image is from the DCGAN paper.

The DCGAN paper also presents other tricks and modifications for training DCGANs like using batch normalization or leaky ReLUs if you’re interested.

Using

G(z)

G(z)

to produce fake images

Let’s pause and appreciate how powerful this formulation of

G(z)

G(z)

is. The DCGAN paper showed how a DCGAN can be trained on a dataset of bedroom images. Then sampling

G(z)

G(z)

will produce the following fake images of what the generator thinks bedrooms looks like. None of these images are in the original dataset!

Generating bedroom images with a DCGAN. This image is from the DCGAN paper.

Also, you can perform vector arithmetic on the

z

z

input space. The following is on a network trained to produce faces.

Face arithmetic with DCGANs. This image is from the DCGAN paper.

[ML-Heavy] Training DCGANs

Now that we have defined

G(z)

G(z)

and have seen how powerful the formulation is, how do we train it? We have a lot of latent variables (or parameters) that we need to find. This is where using adversarial networks comes in.

First let’s define some notation. Let the (unknown) probability distribution of our data be

p

data

pdata

. Also we can interpret

G(z)

G(z)

(where

z∼

p

z

z∼pz

) as drawing samples from a probability distribution, let’s call it the generative probability distribution,

p

g

pg

.

Probability Distribution Notation

Meaning

p

z

pz

The (known, simple) distribution

z

z

goes over

p

data

pdata

The (unknown) distribution over our images. This is where our images are sampled from.

p

g

pg

The generative distribution that the generator

G

G

samples from. We would like for

p

g

=

p

data

pg=pdata

The discriminator network

D(x)

D(x)

takes some image

x

x

on input and returns the probability that the image

x

x

was sampled from

p

data

pdata

. The discriminator should return a value closer to 1 when the image is from

p

data

pdata

and a value closer to 0 when the image is fake, like an image sampled from

p

g

pg

. In DCGANs,

D(x)

D(x)

is a traditional convolutional network.

The discriminator convolutional network. This image is from the inpainting paper.

The goal of training the discriminator

D(x)

D(x)

is:

Maximize D(x) D(x) for every image from the true data distribution x∼p data x∼pdata .

Minimize D(x) D(x) for every image not from the true data distribution x≁p data x≁pdata .

The goal of training the generator

G(z)

G(z)

is to produce samples that fool

D

D

. The output of the generator is an image and can be used as the input to the discriminator. Therefore, the generator wants to to maximize

D(G(z))

D(G(z))

, or equivalently minimize

1−D(G(z))

1

−D(G(z))

because

D

D

is a probability estimate and only ranges between 0 and 1.

As presented in the paper, training adversarial networks is done with the following minimax game. The expectations in the first term go over the samples from the true data distribution and over samples from

p

z

pz

in the second term, which goes over

G(z)∼

p

g

G(z)∼pg

.

min

G

max

D

𝔼

x∼

p

data

logD(x)+

𝔼

z∼

p

z

[log(1−D(G(z)))]

minGmaxDEx∼pdatalog⁡D(x)+Ez∼pz[log⁡(

1

−D(G(z)))]

We will train

D

D

and

G

G

by taking the gradients of this expression with respect to their parameters. We know how to quickly compute every part of this expression. The expectations are approximated in minibatches of size

m,

m,

and the inner maximization can be approximated with

k

k

gradient steps. It turns out

k=1

k=

1

works well for training.

Let

θ

d

θd

be the parameters of the discriminator and

θ

g

θg

be the parameters the generator. The gradients of the loss with respect to

θ

d

θd

and

θ

g

θg

can be computed with backpropagation because

D

D

and

G

G

are defined by well-understood neural network components. Here’s the training algorithm from the GAN paper. Ideally once this is finished,

p

g

=

p

data

pg=pdata

, so

G(z)

G(z)

will be able to produce new samples from

p

data

pdata

.

GAN training algorithm from the GAN paper.

Existing GAN and DCGAN implementations

There are many great GAN and DCGAN implementations on GitHub you can browse:

goodfeli/adversarial: Theano GAN implementation released by the authors of the GAN paper.

tqchen/mxnet-gan: Unofficial MXNet GAN implementation.

Newmu/dcgan_code: Theano DCGAN implementation released by the authors of the DCGAN paper.

soumith/dcgan.torch: Torch DCGAN implementation by one of the authors (Soumith Chintala) of the DCGAN paper.

carpedm20/DCGAN-tensorflow: Unofficial TensorFlow DCGAN implementation.

openai/improved-gan: Code behind OpenAI’s first paper. Extensively modifies carpedm20/DCGAN-tensorflow.

mattya/chainer-DCGAN: Unofficial Chainer DCGAN implementation.

jacobgil/keras-dcgan: Unofficial (and incomplete) Keras DCGAN implementation.

Moving forward, we will build on carpedm20/DCGAN-tensorflow.

[ML-Heavy] DCGANs in TensorFlow

The implementation for this portion is in my bamos/dcgan-completion.tensorflow GitHub repository. I strongly emphasize that the code in this portion is from Taehoon Kim’s carpedm20/DCGAN-tensorflow repository. We’ll use my repository here so that we can easily use the image completion portions in the next section.

The implementation is mostly in a Python class called DCGAN in model.py. It’s helpful to have everything in a class like this so that intermediate states can be saved after training and then loaded for later use.

First let’s define the generator and discriminator architectures. The linear, conv2d_transpose, conv2d, and lrelu functions are defined in ops.py.

def generator(self, z):

   self.z_, self.h0_w, self.h0_b = linear(z, self.gf_dim*8*4*4,

                                          'g_h0_lin', with_w=True)

   self.h0 = tf.reshape(self.z_, [-1, 4, 4, self.gf_dim * 8])

   h0 = tf.nn.relu(self.g_bn0(self.h0))

   self.h1, self.h1_w, self.h1_b = conv2d_transpose(h0,

       [self.batch_size, 8, 8, self.gf_dim*4], name='g_h1', with_w=True)

   h1 = tf.nn.relu(self.g_bn1(self.h1))

   h2, self.h2_w, self.h2_b = conv2d_transpose(h1,

       [self.batch_size, 16, 16, self.gf_dim*2], name='g_h2', with_w=True)

   h2 = tf.nn.relu(self.g_bn2(h2))

   h3, self.h3_w, self.h3_b = conv2d_transpose(h2,

       [self.batch_size, 32, 32, self.gf_dim*1], name='g_h3', with_w=True)

   h3 = tf.nn.relu(self.g_bn3(h3))

   h4, self.h4_w, self.h4_b = conv2d_transpose(h3,

       [self.batch_size, 64, 64, 3], name='g_h4', with_w=True)

   return tf.nn.tanh(h4)

def discriminator(self, image, reuse=False):

   if reuse:

       tf.get_variable_scope().reuse_variables()

   h0 = lrelu(conv2d(image, self.df_dim, name='d_h0_conv'))

   h1 = lrelu(self.d_bn1(conv2d(h0, self.df_dim*2, name='d_h1_conv')))

   h2 = lrelu(self.d_bn2(conv2d(h1, self.df_dim*4, name='d_h2_conv')))

   h3 = lrelu(self.d_bn3(conv2d(h2, self.df_dim*8, name='d_h3_conv')))

   h4 = linear(tf.reshape(h3, [-1, 8192]), 1, 'd_h3_lin')

   return tf.nn.sigmoid(h4), h4

When we’re initializing this class, we’ll use these functions to create the models. We need two versions of the discriminator that shares (or reuses) parameters. One for the minibatch of images from the data distribution and the other for the minibatch of images from the generator.

self.G = self.generator(self.z)

self.D, self.D_logits = self.discriminator(self.images)

self.D_, self.D_logits_ = self.discriminator(self.G, reuse=True)

Next, we’ll define the loss functions. Instead of using the sums, we’ll use the cross entropy between

D

D

’s predictions and what we want them to be because it works better. The discriminator wants the predictions on the “real” data to be all ones and the predictions on the “fake” data from the generator to be all zeros. The generator wants the discriminator’s predictions to be all ones.

self.d_loss_real = tf.reduce_mean(

   tf.nn.sigmoid_cross_entropy_with_logits(self.D_logits,

                                           tf.ones_like(self.D)))

self.d_loss_fake = tf.reduce_mean(

   tf.nn.sigmoid_cross_entropy_with_logits(self.D_logits_,

                                           tf.zeros_like(self.D_)))

self.d_loss = self.d_loss_real + self.d_loss_fake

self.g_loss = tf.reduce_mean(

   tf.nn.sigmoid_cross_entropy_with_logits(self.D_logits_,

                                           tf.ones_like(self.D_)))

Gather the variables for each of the models so they can be trained separately.

t_vars = tf.trainable_variables()

self.d_vars = [var for var in t_vars if 'd_' in var.name]

self.g_vars = [var for var in t_vars if 'g_' in var.name]

Now we’re ready to optimize the parameters and we’ll use ADAM, which is an adaptive non-convex optimization method commonly used in modern deep learning. ADAM is often competitive with SGD and (usually) doesn’t require hand-tuning of the learning rate, momentum, and other hyper-parameters.

d_optim = tf.train.AdamOptimizer(config.learning_rate, beta1=config.beta1) \

                   .minimize(self.d_loss, var_list=self.d_vars)

g_optim = tf.train.AdamOptimizer(config.learning_rate, beta1=config.beta1) \

                   .minimize(self.g_loss, var_list=self.g_vars)

We’re ready to go through our data. In each epoch, we sample some images in a minibatch and run the optimizers to update the networks. Interestingly if

G

G

is only updated once, the discriminator’s loss does not go to zero. Also, I think the additional calls at the end to d_loss_fake and d_loss_real are causing a little bit of unnecessary computation and are redundant because these values are computed as part of d_optim and g_optim. As an exercise in TensorFlow, you can try optimizing this part and send a PR to the original repo. (If you do, ping me and I’ll update it in mine too.)

for epoch in xrange(config.epoch):

   ...

   for idx in xrange(0, batch_idxs):

       batch_images = ...

       batch_z = np.random.uniform(-1, 1, [config.batch_size, self.z_dim]) \

                   .astype(np.float32)

       # Update D network

       _, summary_str = self.sess.run([d_optim, self.d_sum],

           feed_dict={ self.images: batch_images, self.z: batch_z })

       # Update G network

       _, summary_str = self.sess.run([g_optim, self.g_sum],

           feed_dict={ self.z: batch_z })

       # Run g_optim twice to make sure that d_loss does not go to zero

       # (different from paper)

       _, summary_str = self.sess.run([g_optim, self.g_sum],

           feed_dict={ self.z: batch_z })

       errD_fake = self.d_loss_fake.eval({self.z: batch_z})

       errD_real = self.d_loss_real.eval({self.images: batch_images})

       errG = self.g_loss.eval({self.z: batch_z})

That’s it! Of course the full code has a little more book-keeping that you can check out in model.py.

Running DCGAN on your images

If you skipped the last section, but are interested in running some code: The implementation for this portion is in my bamos/dcgan-completion.tensorflow GitHub repository. I strongly emphasize that the code in this portion is from Taehoon Kim’s carpedm20/DCGAN-tensorflow repository. We’ll use my repository here so that we can easily use the image completion portions in the next section. As a warning, if you don’t have a CUDA-enabled GPU, training the network in this portion may be prohibitively slow.

Please message me if the following doesn’t work for you!

First let’s clone my bamos/dcgan-completion.tensorflow and OpenFace repositories. We’ll use OpenFace’s Python-only portions to pre-process images. Don’t worry, you won’t have to install OpenFace’s Torch dependency. Create a new working directory for this and clone the repositories:

git clone https://github.com/cmusatyalab/openface.git

git clone https://github.com/bamos/dcgan-completion.tensorflow.git

Next, install OpenCV and dlib for Python 2. (OpenFace currently uses Python 2, but if you’re interested, I’d be happy if you make it Python 3 compatible and send in a PR mentioning this issue.) These can a little tricky to get set up and I’ve included a few notes on what versions I use and how I install in the OpenFace setup guide. Next, install OpenFace’s Python library so we can preprocess images. If you’re not using a virtual environment, you should use sudo when running setup.py to globally install OpenFace. (If you have trouble setting up this portion, you can also use our OpenFace docker build as described in the OpenFace setup guide.)

cd openface

pip2 install -r requirements.txt

python2 setup.py install

models/get-models.sh

cd ..

Next download a dataset of face images. It doesn’t matter if they have labels or not, we’ll get rid of them. A non-exhaustive list of options are: MS-Celeb-1M, CelebA, CASIA-WebFace, FaceScrub, LFW, and MegaFace. Place the dataset in dcgan-completion.tensorflow/data/your-dataset/raw to indicate it’s the dataset’s raw images.

Now we’ll use OpenFace’s alignment tool to pre-process the images to be 64x64.

./openface/util/align-dlib.py data/dcgan-completion.tensorflow/data/your-dataset/raw align innerEyesAndBottomLip data/dcgan-completion.tensorflow/data/your-dataset/aligned --size 64

And finally we’ll flatten the aligned images directory so that it just contains images and no sub-directories.

cd dcgan-completion.tensorflow/data/your-dataset/aligned

find . -name '*.png' -exec mv {} . \;

find . -type d -empty -delete

cd ../../..

We’re ready to train the DCGAN. After installing TensorFlow, start the training.

./train-dcgan.py --dataset ./data/your-dataset/aligned --epoch 20

You can check what randomly sampled images from the generator look like in the samples directory. I’m training on the CASIA-WebFace and FaceScrub datasets because I had them on hand. After 14 epochs, the samples from mine look like:

Samples from my DCGAN after training for 14 epochs with the combined CASIA-WebFace and FaceScrub dataset.

You can also view the TensorFlow graphs and loss functions with TensorBoard.

tensorboard --logdir ./logs

TensorBoard loss visualizations. Will be updated in real-time when training.

TensorBoard visualization of DCGAN networks.

Step 3: Finding the best fake image for image completion

Image completion with DCGANs

Now that we have a trained discriminator

D(x)

D(x)

and generator

G(z)

G(z)

, how can we use them to complete images? In this section I present the techniques in Raymond Yeh and Chen Chen et al.’s paper “Semantic Image Inpainting with Perceptual and Contextual Losses,” which was just posted on arXiv on July 26, 2016.

To do completion for some image

y

y

, something reasonable that doesn’t work is to maximize

D(y)

D(y)

over the missing pixels. This will result in something that’s neither from the data distribution (

p

data

pdata

) nor the generative distribution (

p

g

pg

). What we want is a reasonable projection of

y

y

onto the generative distribution.

(a): Ideal reconstruction of

y

y

onto the generative distribution (the blue manifold). (b): Failure example of trying to reconstruct

y

y

by only maximizing

D(y)

D(y)

. This image is from the inpainting paper.

[ML-Heavy] Loss function for projecting onto

p

g

pg

To define a reasonable projection, let’s first define some notation for completing images. We use a binary mask

M

M

that has values 0 or 1. A value of 1 represents the parts of the image we want to keep and a value of 0 represents the parts of the image we want to complete. We can now define how to complete an image

y

y

given the binary mask

M

M

. Multiply the elements of

y

y

by the elements of

M

M

. The element-wise product between two matrices is sometimes called the Hadamard product and is represented as

M⊙y

M⊙y

.

M⊙y

M⊙y

gives the original part of the image.

Illustration of a binary mask.

Next, suppose we’ve found an image from the generator

G(

z

̂

)

G(z^)

for some

z

̂

z^

that gives a reasonable reconstruction of the missing portions. The completed pixels

(1−M)⊙G(

z

̂

)

(

1

−M)⊙G(z^)

can be added to the original pixels to create the reconstructed image:

x

reconstructed

=M⊙y+(1−M)⊙G(

z

̂

)

xreconstructed=M⊙y+(

1

−M)⊙G(z^)

Now all we need to do is find some

G(

z

̂

)

G(z^)

that does a good job at completing the image. To find

z

̂

z^

, let’s revisit our goals of recovering contextual and perceptual information from the beginning of this post and pose them in the context of DCGANs. We’ll do this by defining loss functions for an arbitrary

z∼

p

z

z∼pz

. A smaller value of these loss functions means that

z

z

is more suitable for completion than a larger value.

Contextual Loss: To keep the same context as the input image, make sure the known pixel locations in the input image

y

y

are similar to the pixels in

G(z)

G(z)

. We need to penalize

G(z)

G(z)

for not creating a similar image for the pixels that we know about. Formally, we do this by element-wise subtracting the pixels in

y

y

from

G(z)

G(z)

and looking at how much they differ:



contextual

(z)=||M⊙G(z)−M⊙y|

|

1

,

Lcontextual(z)=||M⊙G(z)−M⊙y||

1

,

where

||x|

|

1

=

∑

i

|

x

i

|

||x||

1

=∑i|xi|

is the

ℓ

1

ℓ

1

norm of some vector

x

x

. The

ℓ

2

ℓ

2

norm is another reasonable choice, but the inpainting paper says that the

ℓ

1

ℓ

1

norm works better in practice.

In the ideal case, all of the pixels at known locations are the same between

y

y

and

G(z)

G(z)

. Then

G(z

)

i

−

y

i

=0

G(z)i−yi=

0for the known pixels

i

i

and thus



contextual

(z)=0

Lcontextual(z)=

0.

Perceptual Loss: To recover an image that looks real, let’s make sure the discriminator is properly convinced that the image looks real. We’ll do this with the same criterion used in training the DCGAN:



perceptual

(z)=log(1−D(G(z)))

Lperceptual(z)=log⁡(

1

−D(G(z)))

We’re finally ready to find

z

̂

z^

with a combination of the contextual and perceptual losses:

(z)≡



contextual

(z)+λ



perceptual

(z)

L(z)≡Lcontextual(z)+λLperceptual(z)

z

̂

≡arg

min

z

(z)

z^≡argminzL(z)

where

λ

λ

is a hyper-parameter that controls how import the contextual loss is relative to the perceptual loss. (I use

λ=0.1

λ=

0.1

by default and haven’t played with it too much.) Then as before, the reconstructed image fills in the missing values of

y

y

with

G(

z

̂

)

G(z^)

:

x

reconstructed

=M⊙y+(1−M)⊙G(

z

̂

)

xreconstructed=M⊙y+(

1

−M)⊙G(z^)

The inpainting paper also uses poisson blending (see Chris Traile’s post for an introduction to it) to smooth the reconstructed image.

[ML-Heavy] TensorFlow implementation of image completion with DCGANs

This section presents the changes I’ve added to bamos/dcgan-completion.tensorflow that modifies Taehoon Kim’s carpedm20/DCGAN-tensorflow for image completion.

We can re-use a lot of the existing variables for completion. The only new variable we’ll add is a mask for completion:

self.mask = tf.placeholder(tf.float32, [None] + self.image_shape, name='mask')

We’ll solve

arg

min

z

(z)

argminzL(z)

iteratively with gradient descent with the gradients

∇

z

(z)

∇zL(z)

. TensorFlow’s automatic differentiation can compute this for us once we’ve defined the loss functions! So the entire idea of completion with DCGANs can be implemented by just adding four lines of TensorFlow code to an existing DCGAN implementation. (Of course implementing this also involves some non-TensorFlow code.)

self.contextual_loss = tf.reduce_sum(

   tf.contrib.layers.flatten(

       tf.abs(tf.mul(self.mask, self.G) - tf.mul(self.mask, self.images))), 1)

self.perceptual_loss = self.g_loss

self.complete_loss = self.contextual_loss + self.lam*self.perceptual_loss

self.grad_complete_loss = tf.gradients(self.complete_loss, self.z)

Next, let’s define a mask. I’ve only added one for the center portions of images, but feel free to add something else like a random mask and send in a pull request.

if config.maskType == 'center':

   scale = 0.25

   assert(scale <= 0.5)

   mask = np.ones(self.image_shape)

   l = int(self.image_size*scale)

   u = int(self.image_size*(1.0-scale))

   mask[l:u, l:u, :] = 0.0

For gradient descent, we’ll use projected gradient descent with minibatches and momentum to project

z

z

to be in

[−1,1]

[−

1

,

1

]

.

for idx in xrange(0, batch_idxs):

   batch_images = ...

   batch_mask = np.resize(mask, [self.batch_size] + self.image_shape)

   zhats = np.random.uniform(-1, 1, size=(self.batch_size, self.z_dim))

   v = 0

   for i in xrange(config.nIter):

       fd = {

           self.z: zhats,

           self.mask: batch_mask,

           self.images: batch_images,

       }

       run = [self.complete_loss, self.grad_complete_loss, self.G]

       loss, g, G_imgs = self.sess.run(run, feed_dict=fd)

       v_prev = np.copy(v)

       v = config.momentum*v - config.lr*g[0]

       zhats += -config.momentum * v_prev + (1+config.momentum)*v

       zhats = np.clip(zhats, -1, 1)

Completing your images

Select some images to complete and place them in dcgan-completion.tensorflow/your-test-data/raw. Align them as before as dcgan-completion.tensorflow/your-test-data/aligned. I randomly selected images from the LFW for this. My DCGAN wasn’t trained on any of the identities in the LFW.

You can run the completion on your images with:

./complete.py ./data/your-test-data/aligned/* --outDir outputImages

This will run and periodically output the completed images to --outDir. You can create create a gif from these with ImageMagick:

cd outputImages

convert -delay 10 -loop 0 completed/*.png completion.gif

Final image completions. The centers of these images are being automatically generated. The source code to create this is available here. These are not curated! I selected a random subset of images from the LFW dataset.

Conclusion

Thanks for reading, we made it! In this blog post, we covered one method of completing images that:

Interprets images as being samples from a probability distribution.

Generates fake images.

Finds the best fake image for completion.

My examples were on faces, but DCGANs can be trained on other types of images too. In general, GANs are difficult to train and we don’t yet know how to train them on certain classes of objects, nor on large images. However they’re a promising model and I’m excited to see where GAN research takes us!

Feel free to ping me on Twitter @brandondamos, Github @bamos, or elsewhere if you have any comments or suggestions on this post. Thanks!

DCGAN samples (left) and improved GAN samples (right, not covered in this post) on ImageNet showing that we don’t yet understand how to use GANs on every type of image. This image is from the improved GAN paper.

Citation for this article/project

Please consider citing this project in your publications if it helps your research. The following is a BibTeX and plaintext reference. The BibTeX entry requires the url LaTeX package.

@misc{amos2016image,

   title        = {{Image Completion with Deep Learning in TensorFlow}},

   author       = {Amos, Brandon},

   howpublished = {\url{http://bamos.github.io/2016/08/09/deep-completion}},

   note         = {Accessed: [Insert date here]}

}

Brandon Amos. Image Completion with Deep Learning in TensorFlow.

http://bamos.github.io/2016/08/09/deep-completion.

Accessed: [Insert date here]

Partial bibliography for further reading

Raymond Yeh and Chen Chen et al. “Semantic Image Inpainting with Perceptual and Contextual Losses:” Paper this post was based on.

D. Pathak et al. Context Encoders: Feature Learning by Inpainting at CVPR 2016: Another recent method for inpainting that use similar loss functions and have released code on GitHub at pathak22/context-encoder. This method is less computationally expensive than Yeh and Chen et al. because they use a single forward network pass instead of solving an optimization problem that involves many forward and backward passes.

Ian Goodfellow et al. “Generative Adversarial Nets”

Vincent Dumoulin and Francesco Visin. “A guide to convolution arithmetic for deep learning”

Alec Radford, Luke Metz, and Soumith Chintala. “Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks”

Emily Denton et al. “Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks:” Paper behind the EyeScream Project.

Tim Salimans et al. “Improved Techniques for Training GANs:” OpenAI’s first paper. (Not discussed here.)

Bonus: Incomplete Thoughts on TensorFlow and Torch

As a machine learning researcher, I mostly use numpy, Torch, and TensorFlow in my programs. A few years ago, I used Fortran. I implemented OpenFace as a Python library in numpy that calls into networks trained with Torch. Over the past few months, I’ve been using TensorFlow more seriously and have a few thoughts comparing Torch and TensorFlow. These are non-exhaustive and from my personal experiences as a user.

If I am misunderstanding something here, please message me and I’ll add a correction. Due to the fast-paced nature of these frameworks, it’s easy to not have references to everything.

There are many great collections of tutorials, pre-trained models, and technologies for both Torch and TensorFlow in projects like awesome-torch and awesome-tensorflow.

I haven’t used torchnet, but it seems promising.

Torch’s REPL is very nice and I always have it open when I’m developing in Torch to quickly try out operations. TensorFlow’s InteractiveSession is nice, but I find that trying things out interactively is a little slower since everything has to be defined symbolically and initialized in the session. I much prefer trying quick numpy operations in Python’s REPL over TensorFlow operations.

As with a lot of other programming, error messages in TensorFlow and Torch have their own learning curves. Debugging TensorFlow usually involves reasoning about the symbolic constructions while debugging Torch is more concrete. Sometimes the error messages are confusing for me and I send in an issue only to find out that my error was obvious.

Having Python support is critical for my research. I love the scientific Python programming stack and libraries like matplotlib and cvxpy that don’t have an equivalent in Lua. This is why I wrote OpenFace to use Torch for training the neural network, but Python for everything else. I can easily convert TensorFlow arrays to numpy format and use them with other Python code, but I have to work hard to do this with Torch. When I tried using npy4th, I found a bug (that I haven’t reported, sorry) that caused incorrect data to be saved. Torch’s hdf5 bindings seem to work well and can easily be loaded in Python. And for smaller things, I just manually write out logs to a CSV file. Torch has some equivalents to Python, like gnuplot wrappers for some plotting, but I prefer the Python alternatives. There are some Python Torch wrappers like pytorch or lutorpy that might make this easier, but I haven’t tried them and my impression is that they’re not able to cover every Torch feature that can be done in Lua.

In Torch, it’s easy to do very detailed operations on tensors that can execute on the CPU or GPU. With TensorFlow, GPU operations need to be implemented symbolically. In Torch, it’s easy to drop into native C or CUDA if I need to add calls to a C or CUDA library. For example, I’ve created (CPU-only) Torch wrappers around the gurobi and ecos C optimization libraries. In TensorFlow, dropping into C or CUDA is definitely possible (and easy) on the CPU through numpy conversions, but I’m not sure how I would make a native CUDA call. It’s probably possible, but there are no documentation or examples on this.

TensorFlow (built-in) and Torch’s nngraph package graph constructions are both nice. In my experiences for complex graphs, TensorFlow is able to optimize the computations and executes about twice as fast as Torch. I love nngraph’s visualizations, they’re much clearer than TensorBoard’s in my experiences.

TensorBoard is convenient and okay, but I am currently not using it for a few reasons. (In this post, I only used it because carpedm20/DCGAN-tensorflow uses it.) The plots are not publication-quality and modifications are very limited. For example, it’s (currently) impossible to add a rolling average. The TensorBoard data is stored in a protobuf format and there’s currently no documentation or examples on loading the data in my own script. My current solution is to just write out data to CSV files and load and plot them with another script.

I am not surprised to find bugs or missing features in Torch/Lua code. Here’s my PR removing an incorrect rank check to the LAPACK potrs call. I also had to add a potrs wrapper for CUDA. Sometimes I find minor bugs when using TensorFlow/tflearn, but not as frequently and they’re usually minor.

Automatic differentiation in TensorFlow is nice. I can define my loss with one line of code and then get the gradients with one more line. I haven’t used Torch’s autograd package.

The Torch and TensorFlow communities are great at keeping up with the latest deep learning techniques. If a popular idea is released, Torch and TensorFlow implementations are quickly released.

Batch normalization is easier to use in Torch and in general it’s nice to not worry about explicitly defining all of my trainable variables like in TensorFlow. tflearn makes this a little easier in TensorFlow, but I still prefer Torch’s way.

I am not surprised to learn about useful under-documented features in Torch and TensorFlow by reading through somebody else’s source code on GitHub. I described in my last blog post that I found a under-documented way to reduce Torch model sizes in OpenFace.

I prefer how models can be saved and loaded in Torch by passing the object to a function that serializes it and saves it to a single file on disk. In TensorFlow, saving and loading the graph is still functionally the same, but a little more involved. Loading Torch models on ARM is possible, but tricky. I don’t have experience loading TensorFlow models on ARM.

When using multiple GPUs, setting cutorch.setDevice programmatically in Torch is slightly easier than exporting the CUDA_VISIBLE_DEVICES environment variable in TensorFlow.

TensorFlow’s long startup time is a slight annoyance if I want to quickly debug my code on small examples. In Torch, the startup time is negligible.

In Python, I like overriding the process name for long-running experiments with setproctitle so that I can remember what’s running when I look at the running processes on my GPUs or CPUs. However in Torch/Lua, nobody on the mailing list was able to help me do this. I also asked on a Lua IRC channel and somebody tried to help me, but we weren’t able to figure it out.

Torch’s argcheck has a serious limitation when functions have more than 9 optional arguments. In Python, I’d like to start using mypy for simple argument checking, but I’m not sure how well it works in practice with scientific Python.

I wrote gurobi and ecos wrappers because there weren’t any LP or QP solvers in Torch. To my knowledge there still aren’t any other options.

[auburnfamilynews]

Hugh Freeze resigns from Ole Miss

So, you all have to have already seen the news about Hugh Freeze resigning from the head coaching job at Ole Miss after using a university phone to call an escort service. With everything that was swirling regarding NCAA investigations, it’s odd that this is the bit that did him in. Let’s dive right in, shall we?

It all started with this:

Y sources: Explosive new information has put Ole Miss football coach Hugh Freeze's job in immediate jeopardy. A decision should come soon.

— Pat Forde (@YahooForde) July 20, 2017

With then became this:

ANNOUNCEMENT | Hugh Freeze has resigned effective immediately. Matt Luke interim head coach. Press conference live at 7:30 PM CT on ESPNews.

— Ole Miss Football (@OleMissFB) July 20, 2017

Which was brought about because of this:

Ole Miss football coach Hugh Freeze made call to number tied to escort service http://bit.ly/2uhLF6H via @usatoday

— Dan Wolken (@DanWolken) July 20, 2017

Want more? Don’t cross Houston Nutt, apparently... he’ll work his inane magic to make you pay.

How does this affect Auburn at the moment? Well, in the immediate sense...

Ole Miss will be blown out by Auburn this season but Rebels get a win tonight by interrupting Auburn's #SECNTakeover with a press conference

— Brandon Marcello (@bmarcello) July 21, 2017

Yes, the Ole Miss presser preempted Auburn’s SEC Network Takeover coverage, but things were still back in full swing in time for you to see the Tigers beat Oregon. I believe we missed some of the Georgia game from 2013, so here are those highlights just in case you were craving that this morning.

As for the actual play on the field, obviously nobody knows how Ole Miss will rally around their interim coach, Matt Luke. Luke was a center at Ole Miss from 1995-1998, and he’s spent his time since graduation at Murray State, Tennessee, Duke, and now Ole Miss once again as co-offensive coordinator turned interim head coach. Luke has at least been present for Ole Miss’ wins over Alabama recently, so maybe he can rekindle some of that magic this year. He’s still got Shea Patterson to work with, so he’s got some tools, but I don’t think many people are going to expect much from the Rebels this season. This is a giant distraction, to put it mildly.

And to make things worse, it’s not like Freeze’s absence will burn off the NCAA cloud hanging over Oxford. The kinds of allegations levied against the school are still going to have to be addressed.

How would Faulkner have written what happened today? Probably like chapter 19 of As I Lay Dying. http://bit.ly/2ui2Q8i

— Andy Staples (@Andy_Staples) July 21, 2017

For the final word on this matter (at least today), one of our favorites has the dagger.

God even actual fossil Robert Bentley knew about burners.

— Peggy (@peggyrossmanith) July 21, 2017

Back to Auburn, James Crepea has continued his early scout on some of the Tigers’ opponents... let’s get Mercer, Missouri, and Mississippi State. Ole Miss should be coming up next, but that may take a little while to concoct.

Kam Pettway was named to the Doak Walker Award watch list:

Congratulations to @Uno_Kam on being named to the @DoakWalkerAward Watch List.#WarEagle http://pic.twitter.com/FwN5I4wtgF

— Auburn Football (@AuburnFootball) July 20, 2017

And check out this piece from the Montgomery Advertiser on the capitol city native Daniel Thomas. Obviously, Thomas improved greatly this year, finishing the regular season with a pair of picks in the Iron Bowl. Looks like even more work’s coming his way in 2017.

Montgomery native Daniel Thomas 'playing every position' in secondary at #Auburn (w/videos) http://bit.ly/2uhDOq4 via @mgmadvertiser http://pic.twitter.com/9YUSDoHakK

— Duane Rankin (@DuaneRankin) July 18, 2017

It’s definitely almost football season. War Eagle!

from College and Magnolia http://bit.ly/2uibJyH via IFTTT

[flauntpage]

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks

The first couple weeks of the 2017-18 NBA season have been more fun, unpredictable, and mind-boggling than anyone could've guessed. After Gordon Hayward's injury, the Boston Celtics look like they'll never lose again, Aaron Gordon appears to be a budding All-Star, and Cleveland Cavaliers general manager Koby Altman should probably consider blowing everything up and starting all over (kidding!).

Seriously, though, the season is quickly shaping into an entertaining adventure nobody saw coming: an ongoing drama between belief and skepticism. Between the Little Engine That Could and Small Sample Size Mountain. Let's take a closer look.

1. Chandler Parsons Looks (Relatively) Phenomenal

Heading into this season, expectations surrounding Chandler Parsons—at 29, post-several significant knee surgeries, after a year in which no player in the entire league (except, um, his own teammate Andrew Harrison) shot the ball worse—were lower than they will be for Netflix's inevitable rollout of Stranger Things 7. But instead of hobbling around as a $23 million ball mover, Parsons is one of the most efficient players in the entire league—last night's 0-for-4 outing against the Orlando Magic notwithstanding—and possesses its lowest defensive rating.

(When he's on the court, Memphis performs like a 79-win team! When he's off, only 29.)

Parsons isn't blowing by defenders (unless they're named "Frank Kaminsky"), but has finally rediscovered some confidence in his shot after starting the season with a petrified look on his face every time someone passed him the ball. He's averaging more points per 36 minutes than ever, and has spent nearly all his time at the four (a smart, new development that's partly due to JaMychal Green's ankle injury).

Parsons recorded two dunks in his first 100 minutes after a grand total of three in 674 minutes last season. On one play against the Charlotte Hornets, he grabbed a defensive rebound, leisurely dribbled to the top of the arc, and launched an open three. It clanged off the front iron, but that's still an encouraging level of comfort to see from a guy who was booed by his own fanbase a couple weeks ago.

What does all this mean for the Memphis Grizzlies? Parsons has only logged 19 minutes beside Mike Conley and Marc Gasol, and in that time they were outscored by 14.2 points per 100 possessions. But if they can gel some on an upcoming five-game road trip, and Parsons is able to sustain some of his efficiency in a larger role without suffering any health-related setbacks, there's a very good chance this team can not only qualify for the playoffs, but make some genuine noise once they're there.

2. Big Men and Closeouts

This might seem obvious, but with even more traditional centers stepping behind the three-point line this year, the guys asked to stop them are also drifting towards the perimeter more than they used to. The following qualifies as anecdotal evidence within a small sample size, but according to NBA.com Dwight Howard is contesting 2.7 threes per game this season, up from 1.5 last year. Marc Gasol is at 3.7 three-point contests, and last year he averaged 1.8. Steven Adams contested 2.7 threes last year and now he's at 3.6.

Again, these numbers are circumstantial—reliant on minutes, opponents, and scheme in a tiny sample size—and should be read with a grain of salt. Some centers (like Rudy Gobert and DeMarcus Cousins) haven't seen any uptick at all. But what matters here is the reminder that as NBA offenses continue to evolve, individual defenders are being forced to need to sharpen tools they barely used to need.

Centers who bite at Joel Embiid's pump fake, or wildly race out at Brook Lopez with no plan other than to run him off the line, put pressure on help defenders who're forced to either foul, take a very painful charge, or desert their own assignment and surrender an open look elsewhere.

Sprinting to a dead stop and then trying to laterally stick with a ball-handler is incredibly difficult, but in today's NBA this is what once-plodding seven-footers have to do if they want to stay on the floor.

3. Apologies to Jakob Poeltl

I don't think my opinion on a prospect has ever shifted faster than it has with Jakob Poeltl. It was unclear watching him last year how a seven-footer who can't shoot and doesn't possess leap-off-the-screen athleticism could carve out a meaningful role on a winning team.

This opinion was bad. Poeltl is awesome. Not only is he a putback monster who can control the offensive glass against the right matchup (Toronto's offensive rebound rate is 9.3 percent higher with him in the game), but the 22-year-old has also proven to be an agile pick-and-roll finisher, with touch and strength around the rim. His defense is phenomenal, too, particularly when switching out on the perimeter. Poeltl keeps one hand high to bother the shooter's vision, swivels his hips, and slides step for step.

This is valuable, but thanks to Jonas Valanciunas and Bebe Nogueira, Poeltl's playing time isn't as high as his skills suggest it should be.

4. Philly's Expanding Playbook

It's oh so very early, but according to Synergy Sports, the Philadelphia 76ers boast the NBA's most efficient offense after a timeout. This is a massive leap from last year, when, well, they came in dead last, averaging a measly 0.819 points per possession. Some of this is thanks to Brett Brown's willingness to experiment with the most talented and complementary roster he's ever had, and some is just because said talent is able to savage defenses that aren't as focused as they should be.

Ben Simmons is as perceptive as he is physically imposing; the 21-year-old has already figured out how to make opponents pay when they don't execute as tightly as they should (or when they're simply unable to squeeze the ball out of his hands).

After an Iverson cut towards the left wing, Simmons attacks away from the screen once he notices that Dallas Mavericks big Dwight Powell is hugged up on Amir Johnson instead of in position to ice the pick-and-roll.

The next play starts the same, with Simmons once again opening things up by cutting across the elbow. But instead of Johnson setting a screen, Joel Embiid posts up on the left block while three other Sixers (who're all respectable outside threats) clear out to the weakside. Trevor Ariza isn't in position to force Simmons towards the sideline, so the phenom behaves like a phenom and instead plows into the middle towards an open lane.

These two positive results come off action that isn't especially creative. But Brown is smart enough to realize that sometimes all he has to do is get out of the way. Wind up your franchise player, point him towards a simple two-man action, then let him wreak some havoc. Simmons's ability to read and react at warp speed is one of the many unteachable gifts he already has, and the scheme that can slow him down might not currently exist.

5. Is Ricky Rubio Finally Evolving?

Watch what happens when a defense goes out of its way to prevent Rubio from shooting the ball.

As he spins middle off Gobert's screen, Brandon Ingram leaves Joe Ingles (you know, the guy who made 44.1 percent of his threes last season and is even more accurate this year) to stunt and force a pass. The ball is eventually swung to the opposite corner, where Rodney Hood drills an open look.

This is probably more due to an antsy 20-year-old trying to make a play than a tactical decision from Lakers head coach Luke Walton, but it hints at a reality many thought we'd never see. Rubio is making shots. What's even more impressive than him making 38 percent of his threes (and a completely unsustainable 54 percent of his long twos) is a newfound bravery attached to his shot selection.

Rubio's three-point rate is currently 16.7 percent higher than his career average. Above-the-break treys are still all over the place and he still can't finish at the rim, but a willingness to fire away could change how defenses treat him over the course of the season. Off reputation alone, Rubio's gravity won't ever sniff most of his contemporaries, but an ability to make defenders pay every now and again is significant.

(Also, he has the best hair in the league.)

On Wednesday, Rubio finished with 30 points (three short of his career high) on 17 shots. For just one moment, imagine an alternate reality where these developments are taking place on a Jazz roster that also has a healthy Gordon Hayward and Derrick Favors nearly back to the borderline-All-Star plane where he ascended before injuries weakened his antithetical impact. Is that the second or third-best team in the Western Conference? Does a Rubio, Hood, Hayward, Favors, Gobert lineup make the Warriors sweat?

6. Reminder: Giannis is Huge!

The sight never gets old. In the opening few minutes of Milwaukee's blowout loss against Oklahoma City on Tuesday night, Giannis glided around the floor as a taller, stronger, longer, version of all the various wing defenders employed by the Thunder. It was funny, watching OKC's fundamental identity and nightly advantage look so delicate standing beside the NBA's very own Cloverfield. On the same court as Giannis, Paul George, Andre Roberson, and Jerami Grant looked like raptors flailing around in Jurassic Park's final scene.

7. Reminder: De'Aaron Fox is Fast!

Keep an eye on the shot clock.

8. Is 2017-18 Kemba Walker About to Become 2016-17 Isaiah Thomas?

Meaning, are we in store for a second unexpected leap from a spunky Eastern Conference point guard, one season after it felt like they already spilled out all they had to offer? Earlier this week, Walker ranked third in fourth-quarter scoring (he's now at 10th, with a number that would be top five last season), has never been more efficient from inside or outside the arc, and has damn near doubled his free-throw rate.

Walker has been fantastic inside the paint, and the Charlotte Hornets look deprived of all five senses when he's off the court. This is somewhat due to the fact that they don't currently have a backup point guard, but Charlotte is still an unbelievable 33.6 points per 100 possessions better when he's in there.

There's a jumpy, unpredictability to Walker's game right now. On one recent possession against the Memphis Grizzlies, Walker pushed the ball in transition and nearly penetrated beneath the basket before he decided to pump the brakes and dribble back out to set up the offense. But once he realized no Grizzlies were nearby to escort him to the perimeter, he curled baseline and knocked down a wide open jumper. Splash.

With more pressure to shoulder a heavier load after Nicolas Batum went down in the preseason, Walker is playing with an unseen self-belief that's steadily elevating his game even higher than last year's All-Star campaign showed it could go. Taming a tiger is less complicated than corralling him off a high screen right now. He's a virtual lock to make his second-straight All-Star team.

9. The New York Knicks are Rebounding the Shit out of the Ball

Photo by Wendell Cruz - USA TODAY Sports

Remember when the Knicks were mocked for constructing a roster that essentially barred Kristaps Porzingis from spending any time at center (only three percent of his minutes have been at that position this year, down from 21 percent last year)? Well, even after three-straight wins against the Brooklyn Nets, Cleveland Cavaliers, and Denver Nuggets that took place before they were slapped back to Earth by the Houston Rockets, these personnel decisions probably still weren't the way to go.

But what those personnel choices have done is help New York formulate a fun, possibly sustainable (?) Porzingis + Putbacks identity. With Carmelo Anthony out of the picture, Porzingis has spent the opening chapter of his third season mushrooming into an unguardable beanstock. Only Giannis, Boogie, and Steph Curry are averaging more points than Porzingis. Zero players have a higher usage rate.

Instead of spacing issues caused by the likes of Enes Kanter and Kyle O'Quinn, those two have butchered teams on the glass. The Knicks rank second in offensive rebound rate and third in total rebound rate. While almost every other team around the league is downsizing, New York has firmly positioned their 7'3" franchise player at the four. And, relative to some depressing expectations, it's working!

10. I Can't Wait for the Atlanta Hawks to be Good

If you've happened to catch any recent Hawks game at Philips Arena, you might remember sideline reporter Andre Aldridge posted up at a brand new bar that just opened along the court's baseline. It looks like the most amazing place on Earth.

The team is horrible, but have openly cuddled up beside a full-on rebuild that should (if all goes according to plan) make Philips Arena one of the NBA's most lively atmospheres a few years down the road. Until then, Dominique Wilkins and Bob Rathbun need to broadcast every home game games directly from the bar.

11. Let's Trade Jamal Murray for Kyle Lowry

The likeliness of a trade involving these two players is microscopic—the idea disintegrates if the Toronto Raptors and Denver Nuggets both look like solid playoff teams in late January (Lowry can't be dealt until that month)—so I won't spend too much time rationalizing why I think it should happen.

But it sorta makes sense! Big picture, Toronto has a rapidly progressing core simmering beneath its veteran, All-Star-caliber contributors. The aforementioned Poeltl, rookie OG Anunoby, recently signed Norm Powell, and intriguing rotation players like Delon Wright and Pascal Siakam have the future looking solid.

They're successfully rebuilding on the fly while Lowry, DeMar DeRozan, and Serge Ibaka begin to decline on big-money contracts. Trading (at least) one of those three for valuable assets would punt meaningful playoff contention from 2018-2020, but allow continuity to accelerate within a new, modernized offensive system.

If they can somehow land someone with Murray's upside and turn him into their new franchise player, the Raptors would seamlessly glide from a stagnant also-ran to a promising up-and-comer. Dwane Casey has already relented a bit, playing lineups that feature four or five young pups at the same time.

The main holdup here, besides contractual issues that make matching money a little difficult with these two teams, is Denver's cooperation. Why the hell would they give up on a 20-year-old who defends his position and may own the most invaluable offensive trait in basketball: an ability to knock down pull-up threes at a reliable rate?

Denver is almost an inverse of the Raptors. Both teams are operating on two timelines, but the Nuggets are more clearly loaded to do damage five years from now. Nikola Jokic is 22, Gary Harris just turned 23, and Emmanuel Mudiay (who's made 45.5 percent of his threes this year!) is 21. Common sense says "wait." But Paul Millsap's decision to climb aboard turns maximizing the present into a conversation.

Lowry has been pretty bad this year, but he's still one of the five or six most effective all-around players at his position. Imagine how he'd look next to Jokic and Millsap. How much better would Denver be if he's there this season and next?

Again, a trade like this is extremely complicated and would dramatically shift the direction of two franchises that seem to be content with where they are. But the word impossible doesn't exist in today's NBA.

12. Can Rashad Vaughn Maybe Become a Thing?

Vaughn (who recently said "that's what we lived for" in reference to the McGriddle sandwich) entered his third season with one foot in the league and the other on a banana peel. He logged a grand total of four and a half minutes in Milwaukee's first four games (during which he was trade bait) before draining four threes in an 11-point win against the Hawks.

On Halloween, the team decided not to pick up his fourth-year option, making Vaughn an unrestricted free agent this summer. For a team that has little financial flexibility going forward, completely whiffing on a first-round pick can have painful consequences. Giannis is clearly ready to win now, and the Eastern Conference is begging someone to usurp the Cavaliers.

As Malcolm Brogdon, Jabari Parker, and Khris Middleton each become eligible for a significant pay raise in the next couple summers, the pressure will be on Milwaukee's front office to complement their franchise megastar with a championship-caliber supporting cast before he can flee as a free agent.

On paper, Vaughn is an ideal puzzle piece: a 6'6" three-point threat who may one day be able to reliably knock down threes, make plays when the ball is swung his way, and threaten defenses by pulling up off a dribble hand-off or initiating his own pick-and-roll. Maybe the Bucks believe waiting to see if Vaughn pans out is a waste of everybody's time, especially now that Tony Snell already fills the role he was meant for.

But money issues constrict ways in which Milwaukee can improve from the outside. Internal improvement is key. Vaughn's team option feels negligible now, but giving up on him so soon may come back to haunt this team in one way or another.

13. The Spurs are Perfect Even When They're Not

Even though Patty Mills' game-tying three didn't fall, San Antonio's execution of this elevators action at the end of a recent loss against the Indiana Pacers exemplified why they're the coolest cucumbers around.

Everything about this is ideal...until the ball leaves his fingertips.

14. Jordan Clarkson's Usage Rate is Higher than Anthony Davis, Russell Westbrook, and Just About Everybody Else

To suggest Clarkson has made the most of his reduced playing time is to suggest that Kendrick Lamar sometimes steals the spotlight when he's on other people's songs. In ten fewer minutes than he averaged last year, Clarkson is averaging the same amount of points, knocking down threes at a more accurate clip, posting the highest assist rate of his career, and, generally standing out as a quality contributor off Los Angeles' bench. (He launched six threes in 14 freaking minutes against the Toronto Raptors!)

He's efficient for the very first time despite his usage percentage soaring into the rarified air normally reserved for All-Stars. Some of this is because he's the only shot creator on the floor, often paired with the likes of Corey Brewer, Kyle Kuzma, and Josh Hart. And some of it's because he's been instructed to attack. It's too early to speculate whether this is a breakout campaign or just an early-season surge, but Clarkson's production is flying under the radar in a city that thinks Lonzo Ball is the only player who ever lived.

15. Al Horford is Playing Better Defense Than Everybody Else

The Boston Celtics have the best defense in the NBA because Al Horford is playing like its best defender. When he's off the floor they guard like a bottom-10 unit, but when he's out there, nailed down as a human lighthouse guiding Boston's young pack of swarming athletes everywhere they should go, the Celtics are well-choreographed misery.

Individually, the overwhelming talent Horford has had to corral is beyond impressive: Giannis (twice), Ben Simmons, LaMarcus Aldridge, Kristaps Porzingis, and Kevin Love. All opponents are shooting just 56.2 percent at the rim when Horford is on the floor. When he rests, that number spikes all the way up to 74.6 percent. The difference ranks in the 98th percentile among players at his position, according to Cleaning the Glass).

For the fleeting minority that still scoffs at Horford's occasional humdrum box score, and are fed up with the Ambien-akin side-effects commonly linked to what happens after repeated exposure to negated entry passes, crisp high screens, and perfect execution of myriad pick-and-roll coverages, Horford remains an overpaid waste. Nearly a dozen years of evidence proves they're wrong, and this year he's definitively worth every penny.

Using priceless instincts, flawless habits, and a wingspan that allows him to cover more ground than anyone his size should (only seven players contest more shots every game, per NBA.com), Horford has glued himself inside the all-too-early Defensive Player of the Year (pseudo-MVP?) conversation. He shouldn't leave it anytime soon.

16. Tristan Thompson is a Black Eye on Cleveland's Bloody Face

The Cavaliers have dropped five of their last six games, with all five losses coming up against teams few, if anybody, projected to make the playoffs. Life is rough. But on a team with defensive woes that are as much due to mental indifference as they are physical fragility, Thompson's struggles across the board are particularly worrisome.

Two years ago, the Cavaliers allowed 101.7 points per 100 possessions with Thompson on the floor. This season, his defensive rating is 111.2. His minutes are down, his confidence is low, and his offensive role is non-existent. It's obviously possible for the Cavaliers to bounce back after Isaiah Thomas returns and LeBron James starts to feel like a superhero.

But up until he suffered a calf injury against the Indiana Pacers that will sideline him about a month, Thompson was a non-threat off the ball who launched more long twos than he ever should. If James leaves in free agency this summer, the $36 million Thompson is owed over the next two years turn that contract into one of the league's roughest (from Cleveland's perspective!) agreements.

To be fair, once he's healthy, Thompson's numbers should stabilize once Cleveland works an actual point guard into their rotation. Teams were able to switch James-Thompson pick-and-rolls, and the sliver of opportunity born from that action mainly arrived after a mistake. Here's an example, as miscommunication between Jrue Holiday and Dante Cunningham leads to an easy dunk.

17. Dillon Brooks is Found Money

I wonder how a lucky a front office feels whenever they draft someone 45th overall and then immediately watch him flourish in consequential ways. Is this like finding a $20 bill in your back pocket or hearing your train approach the second you descend onto a subway platform?

The Memphis Grizzlies have had their fair share of first-round blunders, but scoring with guys like Brooks has helped keep this organization afloat, stiff-arming a rebuild further out than it probably should be.

I don't have much to say about Brooks. He seems to be a cagey one-on-one defender, someone who's relentless and difficult to screen. That's nice. He's also committed a bunch of rookie mistakes and isn't really making his threes. But the fact that he's averaging 30 minutes per game on one of the league's most pleasant surprises is telling.

The value of a second-round pick is never more clear than in transcendent figures like Manu Ginobili or Draymond Green, but they still feel like an undervalued commodity. Think about how different the Los Angeles Clippers might look today if they drafted someone like Brooks a few years ago?

Plucking a helpful contributor in the second round takes quite a bit of luck, but some teams have an ability to carve their own more often than others.

18. Your Weekly Reminder that the Golden State Warriors are Unfair

Kevin Durant is shooting 49 percent from behind the three-point line, and his three-point rate has never been higher.

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks published first on http://ift.tt/2pLTmlv

[flauntpage]

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks

The first couple weeks of the 2017-18 NBA season have been more fun, unpredictable, and mind-boggling than anyone could've guessed. After Gordon Hayward's injury, the Boston Celtics look like they'll never lose again, Aaron Gordon appears to be a budding All-Star, and Cleveland Cavaliers general manager Koby Altman should probably consider blowing everything up and starting all over (kidding!).

Seriously, though, the season is quickly shaping into an entertaining adventure nobody saw coming: an ongoing drama between belief and skepticism. Between the Little Engine That Could and Small Sample Size Mountain. Let's take a closer look.

1. Chandler Parsons Looks (Relatively) Phenomenal

Heading into this season, expectations surrounding Chandler Parsons—at 29, post-several significant knee surgeries, after a year in which no player in the entire league (except, um, his own teammate Andrew Harrison) shot the ball worse—were lower than they will be for Netflix's inevitable rollout of Stranger Things 7. But instead of hobbling around as a $23 million ball mover, Parsons is one of the most efficient players in the entire league—last night's 0-for-4 outing against the Orlando Magic notwithstanding—and possesses its lowest defensive rating.

(When he's on the court, Memphis performs like a 79-win team! When he's off, only 29.)

Parsons isn't blowing by defenders (unless they're named "Frank Kaminsky"), but has finally rediscovered some confidence in his shot after starting the season with a petrified look on his face every time someone passed him the ball. He's averaging more points per 36 minutes than ever, and has spent nearly all his time at the four (a smart, new development that's partly due to JaMychal Green's ankle injury).

Parsons recorded two dunks in his first 100 minutes after a grand total of three in 674 minutes last season. On one play against the Charlotte Hornets, he grabbed a defensive rebound, leisurely dribbled to the top of the arc, and launched an open three. It clanged off the front iron, but that's still an encouraging level of comfort to see from a guy who was booed by his own fanbase a couple weeks ago.

What does all this mean for the Memphis Grizzlies? Parsons has only logged 19 minutes beside Mike Conley and Marc Gasol, and in that time they were outscored by 14.2 points per 100 possessions. But if they can gel some on an upcoming five-game road trip, and Parsons is able to sustain some of his efficiency in a larger role without suffering any health-related setbacks, there's a very good chance this team can not only qualify for the playoffs, but make some genuine noise once they're there.

2. Big Men and Closeouts

This might seem obvious, but with even more traditional centers stepping behind the three-point line this year, the guys asked to stop them are also drifting towards the perimeter more than they used to. The following qualifies as anecdotal evidence within a small sample size, but according to NBA.com Dwight Howard is contesting 2.7 threes per game this season, up from 1.5 last year. Marc Gasol is at 3.7 three-point contests, and last year he averaged 1.8. Steven Adams contested 2.7 threes last year and now he's at 3.6.

Again, these numbers are circumstantial—reliant on minutes, opponents, and scheme in a tiny sample size—and should be read with a grain of salt. Some centers (like Rudy Gobert and DeMarcus Cousins) haven't seen any uptick at all. But what matters here is the reminder that as NBA offenses continue to evolve, individual defenders are being forced to need to sharpen tools they barely used to need.

Centers who bite at Joel Embiid's pump fake, or wildly race out at Brook Lopez with no plan other than to run him off the line, put pressure on help defenders who're forced to either foul, take a very painful charge, or desert their own assignment and surrender an open look elsewhere.

Sprinting to a dead stop and then trying to laterally stick with a ball-handler is incredibly difficult, but in today's NBA this is what once-plodding seven-footers have to do if they want to stay on the floor.

3. Apologies to Jakob Poeltl

I don't think my opinion on a prospect has ever shifted faster than it has with Jakob Poeltl. It was unclear watching him last year how a seven-footer who can't shoot and doesn't possess leap-off-the-screen athleticism could carve out a meaningful role on a winning team.

This opinion was bad. Poeltl is awesome. Not only is he a putback monster who can control the offensive glass against the right matchup (Toronto's offensive rebound rate is 9.3 percent higher with him in the game), but the 22-year-old has also proven to be an agile pick-and-roll finisher, with touch and strength around the rim. His defense is phenomenal, too, particularly when switching out on the perimeter. Poeltl keeps one hand high to bother the shooter's vision, swivels his hips, and slides step for step.

This is valuable, but thanks to Jonas Valanciunas and Bebe Nogueira, Poeltl's playing time isn't as high as his skills suggest it should be.

4. Philly's Expanding Playbook

It's oh so very early, but according to Synergy Sports, the Philadelphia 76ers boast the NBA's most efficient offense after a timeout. This is a massive leap from last year, when, well, they came in dead last, averaging a measly 0.819 points per possession. Some of this is thanks to Brett Brown's willingness to experiment with the most talented and complementary roster he's ever had, and some is just because said talent is able to savage defenses that aren't as focused as they should be.

Ben Simmons is as perceptive as he is physically imposing; the 21-year-old has already figured out how to make opponents pay when they don't execute as tightly as they should (or when they're simply unable to squeeze the ball out of his hands).

After an Iverson cut towards the left wing, Simmons attacks away from the screen once he notices that Dallas Mavericks big Dwight Powell is hugged up on Amir Johnson instead of in position to ice the pick-and-roll.

The next play starts the same, with Simmons once again opening things up by cutting across the elbow. But instead of Johnson setting a screen, Joel Embiid posts up on the left block while three other Sixers (who're all respectable outside threats) clear out to the weakside. Trevor Ariza isn't in position to force Simmons towards the sideline, so the phenom behaves like a phenom and instead plows into the middle towards an open lane.

These two positive results come off action that isn't especially creative. But Brown is smart enough to realize that sometimes all he has to do is get out of the way. Wind up your franchise player, point him towards a simple two-man action, then let him wreak some havoc. Simmons's ability to read and react at warp speed is one of the many unteachable gifts he already has, and the scheme that can slow him down might not currently exist.

5. Is Ricky Rubio Finally Evolving?

Watch what happens when a defense goes out of its way to prevent Rubio from shooting the ball.

As he spins middle off Gobert's screen, Brandon Ingram leaves Joe Ingles (you know, the guy who made 44.1 percent of his threes last season and is even more accurate this year) to stunt and force a pass. The ball is eventually swung to the opposite corner, where Rodney Hood drills an open look.

This is probably more due to an antsy 20-year-old trying to make a play than a tactical decision from Lakers head coach Luke Walton, but it hints at a reality many thought we'd never see. Rubio is making shots. What's even more impressive than him making 38 percent of his threes (and a completely unsustainable 54 percent of his long twos) is a newfound bravery attached to his shot selection.

Rubio's three-point rate is currently 16.7 percent higher than his career average. Above-the-break treys are still all over the place and he still can't finish at the rim, but a willingness to fire away could change how defenses treat him over the course of the season. Off reputation alone, Rubio's gravity won't ever sniff most of his contemporaries, but an ability to make defenders pay every now and again is significant.

(Also, he has the best hair in the league.)

On Wednesday, Rubio finished with 30 points (three short of his career high) on 17 shots. For just one moment, imagine an alternate reality where these developments are taking place on a Jazz roster that also has a healthy Gordon Hayward and Derrick Favors nearly back to the borderline-All-Star plane where he ascended before injuries weakened his antithetical impact. Is that the second or third-best team in the Western Conference? Does a Rubio, Hood, Hayward, Favors, Gobert lineup make the Warriors sweat?

6. Reminder: Giannis is Huge!

The sight never gets old. In the opening few minutes of Milwaukee's blowout loss against Oklahoma City on Tuesday night, Giannis glided around the floor as a taller, stronger, longer, version of all the various wing defenders employed by the Thunder. It was funny, watching OKC's fundamental identity and nightly advantage look so delicate standing beside the NBA's very own Cloverfield. On the same court as Giannis, Paul George, Andre Roberson, and Jerami Grant looked like raptors flailing around in Jurassic Park's final scene.

7. Reminder: De'Aaron Fox is Fast!

Keep an eye on the shot clock.

8. Is 2017-18 Kemba Walker About to Become 2016-17 Isaiah Thomas?

Meaning, are we in store for a second unexpected leap from a spunky Eastern Conference point guard, one season after it felt like they already spilled out all they had to offer? Earlier this week, Walker ranked third in fourth-quarter scoring (he's now at 10th, with a number that would be top five last season), has never been more efficient from inside or outside the arc, and has damn near doubled his free-throw rate.

Walker has been fantastic inside the paint, and the Charlotte Hornets look deprived of all five senses when he's off the court. This is somewhat due to the fact that they don't currently have a backup point guard, but Charlotte is still an unbelievable 33.6 points per 100 possessions better when he's in there.

There's a jumpy, unpredictability to Walker's game right now. On one recent possession against the Memphis Grizzlies, Walker pushed the ball in transition and nearly penetrated beneath the basket before he decided to pump the brakes and dribble back out to set up the offense. But once he realized no Grizzlies were nearby to escort him to the perimeter, he curled baseline and knocked down a wide open jumper. Splash.

With more pressure to shoulder a heavier load after Nicolas Batum went down in the preseason, Walker is playing with an unseen self-belief that's steadily elevating his game even higher than last year's All-Star campaign showed it could go. Taming a tiger is less complicated than corralling him off a high screen right now. He's a virtual lock to make his second-straight All-Star team.

9. The New York Knicks are Rebounding the Shit out of the Ball

Photo by Wendell Cruz - USA TODAY Sports

Remember when the Knicks were mocked for constructing a roster that essentially barred Kristaps Porzingis from spending any time at center (only three percent of his minutes have been at that position this year, down from 21 percent last year)? Well, even after three-straight wins against the Brooklyn Nets, Cleveland Cavaliers, and Denver Nuggets that took place before they were slapped back to Earth by the Houston Rockets, these personnel decisions probably still weren't the way to go.

But what those personnel choices have done is help New York formulate a fun, possibly sustainable (?) Porzingis + Putbacks identity. With Carmelo Anthony out of the picture, Porzingis has spent the opening chapter of his third season mushrooming into an unguardable beanstock. Only Giannis, Boogie, and Steph Curry are averaging more points than Porzingis. Zero players have a higher usage rate.

Instead of spacing issues caused by the likes of Enes Kanter and Kyle O'Quinn, those two have butchered teams on the glass. The Knicks rank second in offensive rebound rate and third in total rebound rate. While almost every other team around the league is downsizing, New York has firmly positioned their 7'3" franchise player at the four. And, relative to some depressing expectations, it's working!

10. I Can't Wait for the Atlanta Hawks to be Good

If you've happened to catch any recent Hawks game at Philips Arena, you might remember sideline reporter Andre Aldridge posted up at a brand new bar that just opened along the court's baseline. It looks like the most amazing place on Earth.

The team is horrible, but have openly cuddled up beside a full-on rebuild that should (if all goes according to plan) make Philips Arena one of the NBA's most lively atmospheres a few years down the road. Until then, Dominique Wilkins and Bob Rathbun need to broadcast every home game games directly from the bar.

11. Let's Trade Jamal Murray for Kyle Lowry

The likeliness of a trade involving these two players is microscopic—the idea disintegrates if the Toronto Raptors and Denver Nuggets both look like solid playoff teams in late January (Lowry can't be dealt until that month)—so I won't spend too much time rationalizing why I think it should happen.

But it sorta makes sense! Big picture, Toronto has a rapidly progressing core simmering beneath its veteran, All-Star-caliber contributors. The aforementioned Poeltl, rookie OG Anunoby, recently signed Norm Powell, and intriguing rotation players like Delon Wright and Pascal Siakam have the future looking solid.

They're successfully rebuilding on the fly while Lowry, DeMar DeRozan, and Serge Ibaka begin to decline on big-money contracts. Trading (at least) one of those three for valuable assets would punt meaningful playoff contention from 2018-2020, but allow continuity to accelerate within a new, modernized offensive system.

If they can somehow land someone with Murray's upside and turn him into their new franchise player, the Raptors would seamlessly glide from a stagnant also-ran to a promising up-and-comer. Dwane Casey has already relented a bit, playing lineups that feature four or five young pups at the same time.

The main holdup here, besides contractual issues that make matching money a little difficult with these two teams, is Denver's cooperation. Why the hell would they give up on a 20-year-old who defends his position and may own the most invaluable offensive trait in basketball: an ability to knock down pull-up threes at a reliable rate?

Denver is almost an inverse of the Raptors. Both teams are operating on two timelines, but the Nuggets are more clearly loaded to do damage five years from now. Nikola Jokic is 22, Gary Harris just turned 23, and Emmanuel Mudiay (who's made 45.5 percent of his threes this year!) is 21. Common sense says "wait." But Paul Millsap's decision to climb aboard turns maximizing the present into a conversation.

Lowry has been pretty bad this year, but he's still one of the five or six most effective all-around players at his position. Imagine how he'd look next to Jokic and Millsap. How much better would Denver be if he's there this season and next?

Again, a trade like this is extremely complicated and would dramatically shift the direction of two franchises that seem to be content with where they are. But the word impossible doesn't exist in today's NBA.

12. Can Rashad Vaughn Maybe Become a Thing?

Vaughn (who recently said "that's what we lived for" in reference to the McGriddle sandwich) entered his third season with one foot in the league and the other on a banana peel. He logged a grand total of four and a half minutes in Milwaukee's first four games (during which he was trade bait) before draining four threes in an 11-point win against the Hawks.

On Halloween, the team decided not to pick up his fourth-year option, making Vaughn an unrestricted free agent this summer. For a team that has little financial flexibility going forward, completely whiffing on a first-round pick can have painful consequences. Giannis is clearly ready to win now, and the Eastern Conference is begging someone to usurp the Cavaliers.

As Malcolm Brogdon, Jabari Parker, and Khris Middleton each become eligible for a significant pay raise in the next couple summers, the pressure will be on Milwaukee's front office to complement their franchise megastar with a championship-caliber supporting cast before he can flee as a free agent.

On paper, Vaughn is an ideal puzzle piece: a 6'6" three-point threat who may one day be able to reliably knock down threes, make plays when the ball is swung his way, and threaten defenses by pulling up off a dribble hand-off or initiating his own pick-and-roll. Maybe the Bucks believe waiting to see if Vaughn pans out is a waste of everybody's time, especially now that Tony Snell already fills the role he was meant for.

But money issues constrict ways in which Milwaukee can improve from the outside. Internal improvement is key. Vaughn's team option feels negligible now, but giving up on him so soon may come back to haunt this team in one way or another.

13. The Spurs are Perfect Even When They're Not

Even though Patty Mills' game-tying three didn't fall, San Antonio's execution of this elevators action at the end of a recent loss against the Indiana Pacers exemplified why they're the coolest cucumbers around.

Everything about this is ideal...until the ball leaves his fingertips.

14. Jordan Clarkson's Usage Rate is Higher than Anthony Davis, Russell Westbrook, and Just About Everybody Else

To suggest Clarkson has made the most of his reduced playing time is to suggest that Kendrick Lamar sometimes steals the spotlight when he's on other people's songs. In ten fewer minutes than he averaged last year, Clarkson is averaging the same amount of points, knocking down threes at a more accurate clip, posting the highest assist rate of his career, and, generally standing out as a quality contributor off Los Angeles' bench. (He launched six threes in 14 freaking minutes against the Toronto Raptors!)

He's efficient for the very first time despite his usage percentage soaring into the rarified air normally reserved for All-Stars. Some of this is because he's the only shot creator on the floor, often paired with the likes of Corey Brewer, Kyle Kuzma, and Josh Hart. And some of it's because he's been instructed to attack. It's too early to speculate whether this is a breakout campaign or just an early-season surge, but Clarkson's production is flying under the radar in a city that thinks Lonzo Ball is the only player who ever lived.

15. Al Horford is Playing Better Defense Than Everybody Else

The Boston Celtics have the best defense in the NBA because Al Horford is playing like its best defender. When he's off the floor they guard like a bottom-10 unit, but when he's out there, nailed down as a human lighthouse guiding Boston's young pack of swarming athletes everywhere they should go, the Celtics are well-choreographed misery.

Individually, the overwhelming talent Horford has had to corral is beyond impressive: Giannis (twice), Ben Simmons, LaMarcus Aldridge, Kristaps Porzingis, and Kevin Love. All opponents are shooting just 56.2 percent at the rim when Horford is on the floor. When he rests, that number spikes all the way up to 74.6 percent. The difference ranks in the 98th percentile among players at his position, according to Cleaning the Glass).

For the fleeting minority that still scoffs at Horford's occasional humdrum box score, and are fed up with the Ambien-akin side-effects commonly linked to what happens after repeated exposure to negated entry passes, crisp high screens, and perfect execution of more pick-and-roll coverages, Horford remains an overpaid waste. Nearly a dozen years of evidence proves they're wrong, and this year he's definitively worth every penny.

Using priceless instincts, flawless habits, and a wingspan that allows him to cover more ground than anyone his size should (only seven players contest more shots every game, per NBA.com), Horford has glued himself inside the all-too-early Defensive Player of the Year (pseudo-MVP?) conversation. He shouldn't leave it anytime soon.

16. Tristan Thompson is a Black Eye on Cleveland's Bloody Face

The Cavaliers have dropped five of their last six games, with all five losses coming up against teams few, if anybody, projected to make the playoffs. Life is rough. But on a team with defensive woes that are as much due to mental indifference as they are physical fragility, Thompson's struggles across the board are particularly worrisome.

Two years ago, the Cavaliers allowed 101.7 points per 100 possessions with Thompson on the floor. This season, his defensive rating is 111.2. His minutes are down, his confidence is low, and his offensive role is non-existent. It's obviously possible for the Cavaliers to bounce back after Isaiah Thomas returns and LeBron James starts to feel like a superhero.

But up until he suffered a calf injury against the Indiana Pacers that will sideline him about a month, Thompson was a non-threat off the ball who launched more long twos than he ever should. If James leaves in free agency this summer, the $36 million Thompson is owed over the next two years turn that contract into one of the league's roughest (from Cleveland's perspective!) agreements.

To be fair, once he's healthy, Thompson's numbers should stabilize once Cleveland works an actual point guard into their rotation. Teams were able to switch James-Thompson pick-and-rolls, and the sliver of opportunity born from that action mainly arrived after a mistake. Here's an example, as miscommunication between Jrue Holiday and Dante Cunningham leads to an easy dunk.

17. Dillon Brooks is Found Money

I wonder how a lucky a front office feels whenever they draft someone 45th overall and then immediately watch him flourish in consequential ways. Is this like finding a $20 bill in your back pocket or hearing your train approach the second you descend onto a subway platform?

The Memphis Grizzlies have had their fair share of first-round blunders, but scoring with guys like Brooks has helped keep this organization afloat, stiff-arming a rebuild further out than it probably should be.

I don't have much to say about Brooks. He seems to be a cagey one-on-one defender, someone who's relentless and difficult to screen. That's nice. He's also committed a bunch of rookie mistakes and isn't really making his threes. But the fact that he's averaging 30 minutes per game on one of the league's most pleasant surprises is telling.

The value of a second-round pick is never more clear than in transcendent figures like Manu Ginobili or Draymond Green, but they still feel like an undervalued commodity. Think about how different the Los Angeles Clippers might look today if they drafted someone like Brooks a few years ago?

Plucking a helpful contributor in the second round takes quite a bit of luck, but some teams have an ability to carve their own more often than others.

18. Your Weekly Reminder that the Golden State Warriors are Unfair

Kevin Durant is shooting 49 percent from behind the three-point line, and his three-point rate has never been higher.

The Outlet Pass: Superstar Kemba, God Mode Horford, and the New Look Knicks published first on http://ift.tt/2pLTmlv