Knicks Morning News (2016.09.23)

  • [ESPN] Jeff Van Gundy: Knicks have potential to win 50 games
    (Thursday, September 22, 2016 1:10:01 PM)

    Jeff Van Gundy: Knicks have potential to win 50 games

  • [NYTimes] Bucks Acquire Michael Beasley From Rockets for Tyler Ennis
    (Thursday, September 22, 2016 10:04:03 PM)

    The Bucks have acquired forward Michael Beasley from the Houston Rockets for backup point guard Tyler Ennis in a deal that beefs up Milwaukee’s frontcourt depth.

  • [NYTimes] Wolves Preparing for Life Without Kevin Garnett
    (Thursday, September 22, 2016 6:27:37 AM)

    Less than a week before the Minnesota Timberwolves open training camp, they still have not heard if franchise icon Kevin Garnett wants to follow Tim Duncan and Kobe Bryant into retirement.

  • [NYTimes] Catchings’ Career Ends and Fever Kneel in Anthem Protest
    (Thursday, September 22, 2016 6:24:44 AM)

    The new WNBA playoff format began with a protest and ended with Phoenix and Atlanta advancing.

  • [NYPost] Why Vegas thinks less of the Knicks than you do
    (Thursday, September 22, 2016 9:39:12 AM)

    Las Vegas has taken a cautious viewpoint on the new-look Knicks, who open training camp Monday with a marquee roster, if one tainted by potential health issues. At the premier Westgate Las Vegas Superbook, the NBA over-under victory totals were released Tuesday, and the Knicks were placed at a sobering 38½ wins. By Wednesday, following…

  • [SNY Knicks] Why Maurice Ndour could endear himself to Knicks fans
    (Thursday, September 22, 2016 12:10:06 PM)

    Though his contract is guaranteed, Maurice Ndour will still need to fight for his spot on the team with plenty of intriguing prospects around.

  • 32 replies on “Knicks Morning News (2016.09.23)”

    I think there’s some confusion about win share calculations on this forum. Milo correct me if I’m wrong on anything, but from my understanding:

    It is the case that a team with more wins has more win shares, since a win share is allocated on the basis of victories. But the an individual player’s win shares is predicated upon their contribution to the team’s total number of wins. In fact the calculation for (offensive) win shares doesn’t include team performance or team record at all–it consists in calculating points produced, marginal offense, etc.

    The difference between good JR Smith on the Knicks and JR Smith on the Cavs is that there were other players on the Cavs who contributed more wins, therefore increasing the total number of win shares–but that doesn’t increase or inflate the individual contribution of JR’s win shares. He contributes a static amount, irrespective of how many wins his team has total.

    Or at least that’s my understanding of WS, if anyone more knowledgable than I wants to correct me, do so, but I think this is right at least in broad strokes.

    I think the people saying that players playing on teams with more wins (or less) having unrepresentative WS is reversing the proper order of explanation. They imagine that the total win shares determines how many win shares each player is apportioned–thereby making “more to spread around”–when really it’s the opposite: the team’s total number of wins is determined by each player’s static contribution, agglomerating into their total number of wins.

    Yep, that’s right. On the defensive end there is some scope for a player to pick up credit from a team’s defensive performance that he may not deserve. But JR isn’t having wins gifted to him by the formula on account of Lebron’s offensive genius.

    For anyone that needs to get even more pumped up about Porzingis, the ESPN Sports Science people just released one of their videos about him:

    What’s on that clip? oh nothing. just comparing his agility (with the ball!) favorably to Isaiah Thomas and D’Angelo Russell, blocking shots up to a 12′ high release point, max reach of 12.5 feet, a max vert of 37.5″ ie. better than Blake Griffin.

    Pumped up!!!!!!

    Silky ….after thinking about it last night you are probably right. The whole thing is making my head hurt lol.

    Yeah I saw that Frank. 50-32

    Yep, that’s right. On the defensive end there is some scope for a player to pick up credit from a team’s defensive performance that he may not deserve. But JR isn’t having wins gifted to him by the formula on account of Lebron’s offensive genius.

    Yeah, Lebron’s offensive genius gets JR more open assisted 3 point attempts, and if JR makes those his WS goes up. JR hit open 3’s like you’d have expected JR Smith to do based on his pre-Cavs career.

    However, his DWS was the highest it had been since he was a member of the 54 wins Knicks team. Playing on the awful Knicks team in ’14 he put up .1 DWS and then 1.3 in a little over double the minutes as a Cavalier.

    yeah, there’s no all in one stat that does a really good job with defense, at least to my knowledge

    i figure his increased dws is because of the Cavs being much better on D than our miserable triangle iterations were, right?

    Yeah, defensive WS is pretty much how a team’s defense performs against league average while the player is on the court. So, if a guy plays a majority of his minutes with really good defenders, he will have a good defensive WS, even if he is kind of shit.

    Edit: Although steals, blocks and defensive rebounds help, too!

    I would love to see regression analysis with all of the shot type data we now have. Players like Kawhi, e.g., get so many steals in the backcourt and finish the possession with a quick dunk. Box score metrics will always be “incomplete,” but we can get a better sense of how valuable those defensive “possession conclusion” stats actually are. Right now, we’re best served by calling a steal whatever the team’s offensive PPP average is. Not exactly scientific.

    As for team data, I wonder how well FGA% on 0-3′ and corner threes correlate with team defensive eFG%.

    From last thread:
    KFiNJ, I also noticed that there is no site that translates single-game data into Win Shares or Wins Produced box-score style. It would be very interesting to total up WS or WP and compare with the actual final score. Are there cases where Team A blows out Team B but Team B has a higher aggregate WP or WS score than Team A? If yes, that suggests that something very significant is missing from those formulas. They may still be useful, but no more so than other catch-all stats, like advanced +/- stats or PER.

    Some lol moments from the press conference:

    Jeff Hornacek: “I don’t think we’re going to have injuries.”

    @TommyBeer: Phil, when asked about goals/expectations for the 2016-17 season: “We’re trying to keep it realistic.”

    It’s hard to believe he has a higher max vertical than Griffin. Also, I think he really struggled to penetrate from the perimeter as a rookie (I mean, it’s not really a skill you could ever expect a guy his size to have), so I think the ball-handling drill is a little bit misleading. He is very fast and quick once he gets started, but seems to lack that quick first step.

    But Anthony Davis doesn’t really have a great first step, either.

    He has an AD ceiling and some of his rookie stats are very similar to AD’s (block rate, turnover rate, etc.)

    But a 7-foot-3 player with a massive wingspan and 37.5 inch vert should finish much better inside of 2 feet than KP did last season (.57.6) That and FT rate are where KP needs to make a big jump next season.

    The dude was 20 years old. He’ll figure it out. He’s already aboard the Immortal Train. Next stop: All-Star Game starter (KP v. KAT would be f’n awesome!)

    Some not-s0-lol moments from the press conference:

    Adam Zagoria ?@AdamZagoria 40m40 minutes ago Manhattan, NY
    Phil on Derrick Rose trial: “It’s not something that’s keeping him up at night.”

    Oh good. I was really worried about that

    Adam Zagoria ?@AdamZagoria 43m43 minutes ago Manhattan, NY
    Phil on if they researched the D-Rose case before trade: ” I don’t think we’re going to talk about it. Thanks anyway for the question.”

    While I agree there should be questions asked about Phil’s acquisition of Rose while he case was/is pending, I don’t disagree with Phil for dodging the question. there’s literally nothing he can say that would be taken in a good light.

    I didn’t follow the entire Win Share conversation, but I think I can clear it up quickly if someone can answer this question accurately.

    Lebron James is on a team made up of himself and 14 “little people”. They go 0-82.

    How many Win Shares does Lebron get?

    The answer is Dred and his brothers would combine for about -25 win shares and Lebron would have, say, 15. You could have just imagined 14 Mudiays without dragging Dreds family into this. Note they add to a low total because my example appreciates that NBA analytics like the BOJ are constrained by the zero lower bound.

    Basketball win shares does not technically start with team wins and divide them like in Bill James, but it does have team components that affect the calculation of points produced and defensive rating as mentioned above, so there is an element of gravity between team wins and cum player WS. Note that The DRTG that goes into Win Shares is not on off data but rather whole team data as calculated by Bref.

    But of course this does not mean that the average player benefits from being on a good team, since this assumes a kind of backwards causation. Some players may benefit for various reasons, but that would be zero sum and detract from other players and merely reflect imperfections in WS. In fact diminishing marginal returns means that WS can be a slight detriment on average to players on better teams.

    I think bpm is probably superior to WS because the coefficients were regressed thoughtfully. Over very long periods xrapm probably beats bpm, and xrapm is basically the same and ESPN’S rpm. But they all have a long way to go and the differences between them aren’t too big until you descend down in quality toward PER and wins produced.

    OK next question. Just to clarify further.

    Some team has a point differential that suggests it should have won 41 games. But because of extraordinary good fortune in close games, a few good calls (bad calls) from the refs etc… they wind up winning 51 games. Long term though, the player stats and everything else about the team suggests they are 41 win team.

    Will that 10 game difference in wins due entirely to good fortune impact the individual player Win Shares?

    Agree that there was no good way to answer the question about looking into the allegations, because clearly the true answer is “no, I didn’t at all.” And that’s not a good answer.

    @Stratomatic, to my knowledge the “extra” wins would not be reflected in individual win shares, because the statistic doesn’t start with the actual win total and divide it from there. It attempts to predict/describe it by adding up the individual production of players.

    @18 team wins are not an input in Bref win shares period.

    How accurate is the sum of Win Shares to actual wins?

    Is it reasonable?

    But remember getting a low std error is no great feat when you do things like use team defensive rtg as an input. Similar to berri thinking he solved the puzzle because he regressed ppp plus noise against wins and found a great r squared.

    But remember getting a low std error is no great feat when you do things like use team defensive rtg as an input. Similar to berri thinking he solved the puzzle because he regressed ppp plus noise against wins and found a great r squared.

    Also, getting a close number to reality does not mean you did it right.

    For example, you could use +/- of each player and assign some ‘wins produced’ to it using a linearized version of the Pythagorean Wins, and it would not deviate much from the actual wins. However +/- is a bad statistic to assess the value of the players.

    Wait a minute, naysayers. It’s a civil trial so Rose isn’t exactly going to jail or anything. If he loses the trial it’ll cost him some money and make him play even harder for his next contract.

    Maybe someday we’ll have players who are motivated for reasons other than losing their rape trial

    two thoughts:
    This is a great discussion of win shares.

    I don’t think that Porzingis gets farther off the ground than Griffin. But since he is taller and has long arms, the same height off the ground means he reaches higher in the air. I think the ESPN video showed how high he could reach when he jumped. That could be higher than Griffin can do.

    Doesn’t Rose face a possible NBA suspension, though? I’d imagine that the rape trial has nothing to do with his motivation to do well this year.

    Regardless of the outcome of the trial or settlement, I find it a bit depressing to root for a franchise that again/still employs men who treat women like crap.

    If you’re looking for purity I suggest that you check the monasteries. I root for the Knicks and I will root for Rose as long as he’s here.

    Comments are closed.