Knicks Morning News (2015.04.24)

  • [New York Times] For Some in the N.B.A., Pregame Chapel Can Be a Spiritual Ankle-Taping (Fri, 24 Apr 2015 09:00:12 GMT)

    There are critics of N.B.A. teams’ pregame chapel services, but many people in the league have made the practice an essential part of their pregame ritual.

  • [New York Times] Warriors Erase 20-Point Hole, Beat Pelicans 123-119 in OT (Fri, 24 Apr 2015 06:50:28 GMT)

    Stephen Curry wasn’t about to turn down the music, or put any kind of damper on a giddy locker room celebration that seemed more reminiscent of a series-clinching triumph.

  • [New York Times] Clippers-Spurs Series Giving Lopsided 1st Round Some Drama (Fri, 24 Apr 2015 06:20:49 GMT)

    One of the most memorable opening rounds in NBA playoff history came last season. One of the least memorable is playing out this season.

  • [New York Times] Rose Scores 34, Bulls Hold Off Bucks 113-106 in Double OT (Fri, 24 Apr 2015 04:17:19 GMT)

    Jimmy Butler intercepted the errant pass and darted the other way for a fastbreak basket.

  • [New York Times] Roundup: Bulls and Celtics Move Within a Game of Sweeping Their First-Round Series (Fri, 24 Apr 2015 03:46:25 GMT)

    Chicago and Cleveland each lead their series by three games to none, a deficit from which no N.B.A. team has ever recovered.

  • [New York Times] James Scores 31, Cavs Beat Celts 103-95 for 3-0 Series Lead (Fri, 24 Apr 2015 03:35:38 GMT)

    LeBron James got to the arena early. When he left the Cleveland Cavaliers had a chance to put an early end to the series.

  • [New York Post] What the Knicks think of Tim Hardaway Jr. after his strong finish (Fri, 24 Apr 2015 00:58:49 -0400)

    Tim Hardaway Jr. has been through three team presidents, two coaches and more teammates he'd care to count in his two seasons after becoming the Knicks' 2013 first-round draft pick….

  • Liked it? Take a second to support Mike Kurylo on Patreon!

    Mike Kurylo

    Mike Kurylo is the founder and editor of KnickerBlogger.net. His book on the 2012 Knicks, "We’ll Always Have Linsanity," is on sale now. Follow him on twitter (@KnickerBlogger).

    102 thoughts to “Knicks Morning News (2015.04.24)”

    1. Steph Curry is an active reminder of how we failed to tank at the end of the 2008-2009 season, landing at the 8 spot, one slot behind where Curry was picked. Still sore.

      A cautionary tale about tanking at the end of the season…

    2. The Knicks also refused to tank in 2007-08, winning 3 of their 6, which ended up seeing them finish 1 game better than Minnesota and Memphis (and 3 ahead of Oklahoma City).

      Minnesota drafted third, Oklahoma City drafted fourth, Memphis fifth and the Knicks sixth.

      Russell Westbrook went fourth.
      Kevin Love went fifth.
      Gallo went sixth.

      I loves me some Gallo, but daaaaaaayum.

      So yeah, just fucking tank when you’re fighting for draft position.

    3. @2

      Uhm, yes, but that was only thanks to David Kahn, who selected Jonny Flynn (!) before Steph Curry, saying that Flynn + Rubio could have worked just like Clyde and Earl the Pearl (go figure). Savvy GMing by Minnesota would probably have left us one position short of Jonny Flynn.

      That said, I see your point.

    4. Minnesota drafted third, Oklahoma City drafted fourth, Memphis fifth and the Knicks sixth.

      Russell Westbrook went fourth.
      Kevin Love went fifth.
      Gallo went sixth.

      Also known as “the draft where McHale owned another GM in a trade”.

    5. Farfa –
      To look at your FWP48 list:

      Rajon Rondo (Dallas stint): -0.003
      Robin Lopez: 0.037
      Amir Johnson: 0.052
      Khris Middleton: 0.082
      Cory Joseph: 0.094
      Draymond Green: 0.116
      DeMarre Carroll: 0.123
      Danny Green: 0.145
      Greg Monroe: 0.146
      Rajon Rondo (Boston stint): 0.161 (crazy, huh?)
      Jimmy Butler: 0.224
      Kawhi Leonard: 0.256 (yup, better than LeBron)

      I find the results fascinating. I think most people on the board would not be thrilled with a max to Monroe, including me…but the numbers make a case. (Particularly if his D is seen as adequate, as per the analysis a few people posted a few days ago.) It would mean no Melo at the 4, so that’s bad, but FWP says it wouldn’t be a terrible signing. So if we came away with Monroe and Danny Green/Cory Joseph, we should be OK, which is both realistic and surprising. Add a top draft pick, and things aren’t as bad as I feared.

      And a max to Middleton or Draymond, at least according to FWP, doesn’t make as much sense as I expected.

      But it does show how hugely valuable Jimmy Butler and Kawhi are, which everyone (except Capn Luke) believed.

    6. I think Monroe would fit in pretty well on the offensive end. My worry is at the other end of the floor. With that being said, someone i forgot who posted Monroe’s defensive #s a couple of days ago and they ok, no where close to as bad as I figured them to be. I’m torn, I still think the Knicks should stay away but the offensive/defensive #s say otherwise. Eh not to worry, I’m sure Phil has scouting Monroe et al all season and has it all figured out.

    7. And a max to Middleton or Draymond, at least according to FWP, doesn’t make as much sense as I expected.

      To be honest my (very raw and imprecise) stat doesn’t say that Draymond doesn’t deserve the max. It says that on the “volume” end guys like DeMarre Carroll and Danny Green have had a much better season.

      I’d say that my “stat” should be seen as a tool to determine the order in which players should be praised for their contributions on the offensive end.

      What I mean is that there is no proportion between numbers (so it’s not like Kawhi has produced more than three times Middleton’s production); at the same time, it seems to paint an apparently accurate picture of how much players have been valuable on the offensive end for their teams (and as a whole). As a comparative stat, it seems to be working.

    8. Count me in as one of the few people who is pretty pro Greg Monroe. He’s only 25 years old and has had a pretty productive career so far. He is decent on D, a great rebounder and scorer. I guess a lot of it depends on who we draft. But if we are able to land Towns…a front court of Towns, Monroe and Melo would be pretty sweet. They’d be a very strong rebounding front court and defensively I think they would pass muster. Monroe and Melo can both score. Also, his “max” is a mini max, so we’d still have like 12 million or so left over in cap space to sign another player or two. I think its not a bad start.

    9. So I’d look at that “stat” to determine if a certain player can help our team on the offensive side while shouldering a significant load. Ideally, I think you should have at least one high WP-high FWP guys on your most played lineup, plus at least three high WP-low FWP guys.

      That said, I won’t bore this board anymore with this stat, this is not about tooting my own horn. It was just a fun exercise to tinker with numbers to give a different interpretation of some of the WP48 results.

    10. I’m ambivalent about Greg Monroe, but that’s only because we already have Melo on our team.

    11. #9 – yeah, that’s starting to look like our best-case scenario for this year:

      Towns
      Monroe
      Melo
      Galloway
      CJoseph

      Not a dream team, but a core to build on, and we can add one more max player the following year. So 2016-2017 might actually be a good year for the Knicks.

      …if we got Towns. If Melo doesn’t fall apart. If Phil doesn’t live down to Luke’s expectations and sign Rondo or something.

    12. “It would mean no Melo at the 4, so that’s bad,”
      I’m not sure if this is necessarily true. There are 96 minutes of 4/5 time to split every game. If Monroe and Towns (hopefully) combine for 62 minutes, that would leave 34 minutes of 4/5 time that someone would have to fill. Maybe Melo could play, say, 20 minutes at the 4 when Monroe is on the bench and Towns is at the 5, or when Towns is on the bench and Monroe is at the 5. Melo at 18-20 minutes at the 4 and 15 or so at the 3 when both Towns and Monroe are in the game might work (although I haven’t spent any time figuring out substitution patterns, so maybe this won’t work at all).

    13. is it bad that i am starting to lean towards Russell. Boy the draft is such a friggin crap shoot glad its not me picking

    14. is it bad that i am starting to lean towards Russell. Boy the draft is such a friggin crap shoot glad its not me picking

      Russell is, like, the world’s greatest consolation prize if they end up #3. He’s going to be a great pro player. But I’d still go Okafor and Towns over him.

    15. I’m still firmly in Towns’ camp. Our roster is so barren that we can not pass (if we have the opportunity) on the chance to draft a big to anchor our paint for the next seven years. I too have the feeling that Russell is going to be a super pro, but no way we don’t draft one between Towns and Okafor if they’re available.

    16. I might go for Russell over Okafor. I’m just not sold on someone who MIGHT be able to learn to defend. I’d almost prefer Cauley-Stein over Okafor. Almost.

    17. Owen, FWP is Farfa’s attempt to adjust for usage. As he explained it:

      We could tweak it a little by subtracting .100 by the WP48 metric (setting the bar at average – average players get a 0) and multiplying by USG%/20 (in doing so, accounting for the offensive burden a guy has to carry in a team).

      A few examples for this FWP48 (Farfa Wins Produced per 48 minutes). The average player should have a score of .000. The “high volume/low efficiency” should have a highly negative score. The “low volume/low efficiency” should have a slightly negative score. The “low volume/high efficiency should have a score around .150. The “high volume/high efficiency should have a highly positive score.)

      The results were interesting, both applied to random players and the Knicks roster. Still not great for good defensive players, but no public metric is. (Which means that Draymond Green, who doesn’t rate quite as highly in FWP, still may be a max player because his D is so good.)

    18. I embraced D. Fisher from the get go as the knicks coach. And I’ve given him a bye for this season, though it was starting to become clear that he may be part of the problem.
      UNTIL THE LAST THREE GAMES!
      At the tail end of this abortion of a season, his selfish desire to notch every win he could at the expense of the long term benefit to the organization – and gloat about it – was a turning point for me. Of course, it was all blessed by his great mentor, BCT.
      I hope the knicks get the 1st pick in the draft, but if they get the 5th, they surely earned it.

    19. I think you have to consider giving him a bye for the last three games too. In Jackson’s recent interview he said that he had consistently told coaches and players to try and win every game, so you have to consider the possibility that Fisher was following orders

    20. Yeah, agreed, he was definitely following orders. If they told him to tank, it’s not like he would pull an audible and just do what he wanted to.

    21. It’s just the latest jackass move by Phil. He’s building a nice portfolio of them.

    22. Is Phil going to say in an interview that he told the team to lose? I’m pretty sure he’d get a hefty fine for that. I really hope he’s astute enough that behind the scenes he was telling Fish to not work too hard at winning those last couple games and that things just didn’t work out (sometimes you do end up winning even when you’re trying to lose). No real way for us to know from the outside though.

    23. “Is Phil going to say in an interview that he told the team to lose? I’m pretty sure he’d get a hefty fine for that. I really hope he’s astute enough that behind the scenes he was telling Fish to not work too hard at winning those last couple games and that things just didn’t work out (sometimes you do end up winning even when you’re trying to lose). No real way for us to know from the outside though.”

      But like the others said, its obvious he didnt. Fisher wouldnt buck Jackson like that.

    24. Im sorry, Not understanding what you wanted Fisher to do?

      Our Starting lineup in the last 3 games was Galloway,Hardaway, Cole, Thomas and Smith who played 36 Min against the hawks? Pretty much the same sorry ass lineup that we have put out the first 79 games beforehand? What did you want Fisher to do?

    25. I now realize that you all just want to see some stats, with no empirical validity, that fascinate you. Be on the lookout for my very first original piece of NBA stats work, Stats Generated Against Replacement Stats. I basically took some other stats that had some weird outliers and I divided them by random prime numbers until they matched up with what I had subjectively determined a metric’s accuracy was. Then I flipped the spreadsheet upside down and changed the font to Wingdings.

      My conclusion is that Bargnani is better than every stat would suggest due to his role on offense never really being dialed in to his strengths, and also he shouldn’t be penalized for being a huge zero on defense because his coaches never really should allow his teams to play a single possession with him on the floor.

    26. Im sorry, Not understanding what you wanted Fisher to do?

      Our Starting lineup in the last 3 games was Galloway,Hardaway, Cole, Thomas and Smith who played 36 Min against the hawks? Pretty much the same sorry ass lineup that we have put out the first 79 games beforehand? What did you want Fisher to do?

      There’s a lot of things you can do. You need only watch certain Philly games to see how a smart coach can take a sorry-ass roster and still amplify it’s chances of losing a specific game if he so desires.

      Have THJ play PG for 48 minutes because “he is a big part of our future and we really want him to work on his ball-handling and play-making”. Same thing but with Early. When you’re winning, go full court press to speed up the game and give your opponent more opportunities. When you’re losing tell your PG to walk it up and call a set on every possession. Try out an ultra-big lineup with 3 centers because you “want to see if Amundson can guard 3s”.

      That’s 2 minutes worth of ideas, and only includes ones that wouldn’t be called out by any but the most astute fans as suspicious in the least. If you’re willing to be more aggressive you can go way further (“Hey Cole we really want you to work on your 3-point shot tonight”). If a coach can help a team win a game he can help them lose one as well.

    27. Oh and I should have mentioned the obvious: If any of those things is working (in the winning games sense), STOP DOING IT! I think it was the Hawks game where Galloway was going nuts: “Hey Langston, great first half tonight, you’re a big part of what we have planned for next year and we don’t want to risk any kind of injury in the 2nd half. Why don’t you grab an early shower and let’s see what Ledo can do?”

      I mean this stuff isn’t rocket science. Losing is supposed to be easier than winning.

    28. Yeah, again, an NBA coach did not get into any trouble at all by having Mark Madsen, who never played at that point in his career, not only play a lot of minutes, but, after not taking a single three all season, take seven in a nine-minute span in the final game of the season in a game his team needed to lose to keep their draft pick. He missed all seven. They lost the game. There were no fines or anything.

      The Warriors did the same thing a couple of years back to protect their draft pick. They ended up getting Harrison Barnes out of it. I don’t think they regret doing that, either.

    29. It’s gonna be awesome when we draft Emmanuel Mudiay, then sign Lamarcus Aldridge. We’re gonna score so many pointz.

    30. “It’s gonna be awesome when we draft Emmanuel Mudiay, then sign Lamarcus Aldridge”

      You guys are killing me putting this stuff into the universe. LOL

    31. Yeah, again, an NBA coach did not get into any trouble at all by having Mark Madsen, who had not taken a three all season, take seven in a nine-minute span in the final game of the season in a game his team needed to lose to keep their draft pick. He missed all seven. They lost the game. There were no fines or anything.

      The Warriors did the same thing a couple of years back to protect their draft pick. They ended up getting Harrison Barnes out of it. I don’t think they regret doing that, either.

      Exactly! I think given all the extra attention that tanking is getting these days such a blatant example would probably land you some serious media scorn and maybe even some league throat clearing, but ultimately what are they going to do? Tell you you’re not allowed to have Cole Aldrich shoot 3s? It’s development!

    32. Jowles – are you mocking Farfa’s model? Because that’s all it is, a model. It seems a pretty rational model that adjusts for usage, with the specific goal of recontextualizing the outliers on offense, like Melo and Chandler. It’s no more bullshit than any other model (except adjusted plus-minus, which with no insight to the formula is totally useless). And the results are interesting…including that Bargnani continues to be abysmally bad.

      So what if it’s built on another model? I mean, the stats we have publicly available are crap. The models are inadequate, some more than others. Since a model is just a pile of formulas that try to take (crappy) statistics and turn them into an explanation of what’s happening, what’s the issue? Or is it that you just like to mock the efforts of others instead of analyzing why their mathematical adjustments are poorly conceived?

    33. @41
      I think the objection is that WP is built on real data, and is already a model that actually has a pretty damn good fit to wins. The relation between USG% and efficiency is already built into the model. Adding another variable that is based on an assumption rather than on data doesn’t make the model more accurate. In fact, it makes it less so.

    34. Actually, I think WP explicitly takes out usage. For example, it calculates offensive efficiency by dividing points scored by possessions used. This isn’t bad, it just makes it different from FWP. Statistics are used to determine what happened, not to point out causes, and you want to use different ones for different purposes. For example batting average tells something different about a player than on base percentage does. By looking at the two statistics together you get a fuller picture of a player than by using one alone. Farfa wanted to sort out players who were efficient but didn’t get used often from ones who got used a lot and were efficient. His statistic does a good job of this and probably reflects NBA player valuations better than WP (even if it doesn’t reflect your valuation better than WP).

    35. The #1 player in WP48 this season was Steph Curry. It’s because he took, and made, a lot of baskets. His high usage is why he has a lot of wins produced.

    36. I thought Farfa was forwarding his FWP as half joking speculation–he may still be, but the fact that people are taking it seriously is crazy. You don’t just randomly come up with the numbers to input in a model and assume they’re appropriate because the output fits your preconceived notions of player value. That’s not a model, that’s an intuition pump.

    37. I have nothing against Farfa, who apparently and inexplicably appreciates my favor, but calling that a model worthy of consideration is totally and utterly insane.

    38. Definition of batshit stupid inane talk — any opinion with which Jowles disagrees. It was batshit talk the other day to point out that Chandler’s one and only unique skill — hitting shots at the rim at 68% for his career — isn’t all that unique. You know who hit shots at the rim at 68% or better this year? Virtually everyone on the Warriors’ roster (including guys like Curry and Iguodala). So, I guess Chandler’s real elite offensive “skill” is refusing to shoot anything but a dunk or layup, regardless of whether his team is struggling to score and could use some points from their unstoppable offensive force (who manages to pour in 20 points once every 25 games or so on average).

    39. Here’s a question for you — everyone here agrees that max contracts should only be given to superstars — guys who really make a difference. If Chandler were a 30 year old free agent, would you give him a 4 year $100 million contract?

    40. If Chandler were a 30 year old free agent, would you give him a 4 year $100 million contract?

      First, I’d take that money and try to give it to DeAndre Jordan. Then, it would probably depend. If I thought my team was a Tyson away from competing for a title for the next two years? Yeah, I would. As long as he’s healthy, Tyson is going to come close to earning that money. But his health risk (and obviously, if he was 30 we wouldn’t know how he would decline) and age would keep me from making that bet if I didn’t already have a good team.

    41. Just to say that in the very moment I decided to call that stat FWP it should have been clear I had no intention to make a serious case at accuracy. I was just curious to see the results.

      That said, the results are pretty interesting, and as I wrote yesterday the real game changer for advanced stats like WP48 would be to know the curve associated to marginal WP48. What happens if Player A increases his usage? What if Player B decreases his? If that curve is a parable, where’s the usage that maximizes that stat? On this, I’m dead serious. You then would have to take expected WP(usage) in consideration when building and/or coaching a team.

    42. Also, Jowles, why do you say “inexplicably”? You are a very smart guy, and game recognizes game :-)

    43. And another thing (sorry, I’ll be quiet after this). The numbers I introduced in my WP spoof are not random. You have to subtract .100 so that a super high usage doesn’t infilate a poor WP48 score (think a 30 USG% with a .075 WP48 against a 15 USG% with a .150; it’s quite clear the first one is a pretty bad chucker, while the second is a capable if maybe limited player. If you multiply those numbers, though, you get the same result) and you divide USG% by 20 to normalize it for the theoretical average USG%.

      It’s not a model and it doesn’t want to be one, but the numbers are not that random.

    44. What happens if Player A increases his usage? What if Player B decreases his?

      That depends on what happens to their shots. There’s been some research done on what happens to efficiency as usage increases/decreases that you should be able to find with a bit of googling, but that isn’t part of wins produced.

    45. These two teams are going to combine for 80 points in the first quarter. That’s insane. Good O or bad D?

    46. Is it me or does Hubbie says the same thing about Tyson every game he calls, “okay, once again Tyson Chandler refuses to step up”, when referring to his lack of rim protection?

    47. Ok Aminu is actually legit which I knew but what’s his contract gonna look like this offseason? He’s a huge value at the minimum which is what the Mavs have him at iirc.

    48. How much do Dallas miss having Wright instead of Stat? The eye test says a lot but i haven’t looked at the #s

    49. That…wow…that was an awful final possession for Dallas. I can’t believe that’s anywhere close to what Carlisle had in mind.

    50. Gortat, or if you’d prefer “The Polish Tyson Chandler,” is having a nice game. 24 points on 15 shots, 11 boards and 3 blocks.

    51. People need to recognize Kawhi is one of the best players in the NBA. Kid is a force.

      Maybe one of the best role players in the league. FWP has him as average. Definitely not a Melo. A product of his system, his coach and his teammates, not his own play.

    52. Piece with 3 huge 3’s in the final minutes. As a Knick fan, he is infuriating. But as a fan of the game, the man has ice in the veins.

    53. Sort of crazy that Austin Rivers, Glen Davis, and Hedo Turkoglu are getting playoff minutes….

      And they are +’s for the night…

    54. Finals MVP. Defensive player of the year. 23 years old. 27 points on 15 shots tonight. This dude is not worth 15 million a year?

    55. Finals MVP. Defensive player of the year. 23 years old. 27 points on 15 shots tonight. This dude is not worth 15 million a year?

      DRed=Mark Jackson?

    56. How many guys in the NBA would you not trade for Kawhi? AD. Curry. Whiteside if he’s for real? Gobert if he’s for real? (they’re even cheaper) That’s pretty much all I got. Lebron is too old and expensive. KD has injury problems.

    57. The Spurs really are set up for a quick rebuild when the big three retire. They’re going to have a huge amount of money when the cap balloons and guys retire, and they already have one of the best young players in the world.
      Farfa, your stat may be made up and not have a lot of statistical meaning, but I’d still put it miles above ESPN’s RPM. At least I know how FWP is calculated.
      As for the questions on usage and WP, if you score efficiently, then your WP will increase with increased USG. It rewards made baskets more than attempts, for the most part. WP does have some quarks that I’m not too sure what to make of, but that’s mostly because it has half a dozen adjustments that don’t necessarily make logical or statistical sense to me. But maybe I’m just dense.

    58. How many guys in the NBA would you not trade for Kawhi? AD. Curry. Whiteside if he’s for real? Gobert if he’s for real? (they’re even cheaper) That’s pretty much all I got. Lebron is too old and expensive. KD has injury problems

      will/could Giannis make this list

    59. The Spurs really are set up for a quick rebuild when the big three retire. They’re going to have a huge amount of money when the cap balloons and guys retire, and they already have one of the best young players in the world.

      While basically true, they’re going to find it a bit tricky this season, as the cap isn’t going up that much this year and they’re going to have to replace a lot of guys all at once. Luckily, Leonard’s cap hold will be low, but still, beyond that, they’re not going to have that much cap room, especially since they really need to get a replacement for Tim Duncan and that will likely not be cheap.

    60. I now realize that you all just want to see some stats, with no empirical validity, that fascinate you.

      THCJ,

      Look at this way. We can accept a model that we know handles the usage issue incorrectly (like Wins Produced, PER, etc..), or we can take what we know Wins Produced does well and tweak in a highly subjective way to produce results that make more sense in the real world.

      It would be much better if Berri would accept that his original research on usage was wrong because he didn’t control for everything and fix Wins Produced. He eventually fixed it for diminishing returns and defensive rebounds after a lot of criticism, but so far he hasn’t been inclined to do so for the extremes of usage. So we are stuck with a subjective adjustment that makes more sense even if not perfect.

      I would rather be approximately right than precisely wrong!

    61. Holy shit this is a funny fight. The WP defenders, thinking they are wielding an ex ante attempt to regress individual boxscore contributions to wins, make fun of the model polluters who (yes, absurdly) tweak the the model to fit their preferred preconception. Yet WP itself is no better than these arbitrary adjustments. The “explanatory” regression in the initial WP model is nothing more than a regression between team statistics and wins and provides absolutely no support for the basic player-level weights assigned in the WP formula. Do you know how many combinations of player boxscore weights can come to equally overwhelming R-squareds when running team level boxscores against wins? It is snake oil. The incremental alterations of WP from this entirely flawed premise (“these weights explain 95% of wins!”) — changing the rebound value, estimating the assist value, even the positional adjustment — that claim to be ex ante data-based adjustment are nothing more than the same ex post and ad hoc alterations of a nonsensical formula as FWP.

      The original WP formulation relied on an obviously flawed “possession” definition that works at the team level but not at the player level. As a result, it weighted (for example) a player rebound as being equal in value to a negative turnover. This is obviously a mistake if you do any thinking, but it doesn’t pollute the team level regression because they naturally offset in the definition of possessions. The formula then has to pretend to make empirically driven “adjustments” to be less silly which is no different than adding a random Farfa variable.

      The history of boxscore stats goes like this.
      (1) People overrate scoring and underrate efficiency & probably the cost of turnovers
      (2) Every advanced metric except stupid PER mostly just add value by capturing this effect.
      (3) Berri hyped his multivariate regression which was worse than everything but PER but did mostly capture the efficiency…

    62. Boy, I never thought someone would be arguing so much about this silly thing.
      Where did I say I was creating something better than WP? Where did I say I had a preconception against Tyson Chandler or players like him? Actually, I think WP does a great job at measuring the overall contributions of any player. I had no preconception whatsoever. I just proposed a thought exercise. By the way, FWP (again, lol at the name itself! do you think I’m taking myself that seriously?) essentially gives the same results of WP. Jowles, you didn’t read correctly: Kawhi is a superstar in both cases. My thought was that tweaking WP48 I could make more obvious some hidden considerations.

      Look, I’ll say it clearly and publicly (no sarcasm whatsoever): WP48 is miles and miles better and tremendously more accurate (and legit) than my poor thing. Can we move on? Really.

    63. Hey Cock – as Farfa said, you’re not even reading correctly, because his FWP had Kawhi as a superstar. I found it interesting for exactly the reasons ptmilo listed: it is an alteration of a FORMULA based on statistics (which, by the way, are themselves, at this current moment, somewhat unreliable). It is not changing the data itself; it is a modification of an INTERPRETATION. A deeply flawed interpretation.

      Do you know how many combinations of player boxscore weights can come to equally overwhelming R-squareds when running team level boxscores against wins? It is snake oil. The incremental alterations of WP from this entirely flawed premise (“these weights explain 95% of wins!”) — changing the rebound value, estimating the assist value, even the positional adjustment — that claim to be ex ante data-based adjustment are nothing more than the same ex post and ad hoc alterations of a nonsensical formula as FWP.

      Exactly. I would disagree that Farfa’s modification is “nonsensical,” as it is an attempt to deal with the extremes of usage that Berri’s formula doesn’t deal with based on the reality that there are five players on the floor who on equal usage would each contribute 20%. But either way, it’s a thought experiment, and though only half-serious about it (as Farfa himself was only half-serious), I thought the conclusions were surprisingly compelling. Kawhi, LeBron, Harden, Curry, and other stars are still stars. But players with extremely high usage and moderate efficiency – Melo – are shown to be only slightly above average, and players with extremely low usage and high efficiency – Chandler – are shown to be valuable, but not AS valuable. That makes a fuckload of sense to me, and since Berri’s formula is arbitrary bullshit, I’m pretty OK with an adjustment to that arbitrary bullshit that correlates to observable results. (Chandler’s presence not affecting a team’s ORT nearly as much as, say, Harden’s or Lebron’s…

    64. Here again were Farfa’s results, which, again, interested me not because they are rigorous formulae based on precisely measured SportsVu data, but because they take a model based on publicly available data and make it less incredible:

      Steph Curry gets a FWP48 score of .406
      LeBron James gets a FWP48 score of .239
      DeAndre Jordan gets a FWP48 score of .190
      Tyson Chandler gets a FWP48 score of .153

      Sounds reasonable, right?

      Let’s look at the “scorers”, instead.

      Kobe Bryant gets a FWP48 score of -.190
      Carmelo Anthony gets a FWP48 score of .031
      Kyrie Irving gets a FWP48 score of .123
      James Harden gets a FWP48 of .347

      Those results do seem much more reasonable than Berri’s. Put another way, I like FWP’s “batshit crazy” results a lot more than Berri’s. So don’t have fun with it, or don’t take it apart with precision as ptmilo does, but try not to be a kneejerk bleach-prescribing ass who’s as besotted with his pseudoscience as various proponents of the “eye-test” are.

    65. I dunno, either low-usage, good-rebounding big men like Jeff Adrien and Furkan Aldemir (?) represent the most severe market inefficiency in sports history or something is off with WP. That type of player seems to do very, very well with the metric.

    66. Well, that thing JK47 said is exactly why it is so hard trusting 100% WP48. In fact, I think WP48 does a marvelous job at highlighting the bad players, but has a strange scale at the top end. And, worst of all, it seems like its creators and most frequent users are all like “well fuck it, if it says so Furkan Aldemir has to be a top 30 NBA player”. Again, I still think WP48 is the best thing out there, but this doesn’t mean we should accept every WP48 score as if it was made of solid gold. Guys thought the sun orbited around the earth for many years, or did they?

    Comments are closed.