Poll: Jackson Year 2 Record

From the New York Times:

Pressed on what he meant, Jackson said he hoped the Knicks could make the playoffs — which he also predicted last fall before blowing up an underachieving roster in midseason — though not as the Nets did this season, with a losing record.

“I hope it’s not at the extent that well, we just made the playoffs and we’re not 35 and 47,” he said. “I want to see us having a winning record, and that’s a big jump to make.”

On Pardon the Interruption, Tony and Mike brought up Jackson’s belief that the Knicks would have a winning record in 2016. With his psychic costume on, Kornheiser predicted the Knicks would sign Kevin Love and win 42-45 games. Allowing for Tony to keep his fantasy of the Knicks signing Kevin Love, Wilbon asked how many seasons prior to Cleveland did Love have a winning record (answer: none). Then he predicted a 38-40 win season for New York.

So I was curious where our readers stand.

Will the Knicks have a winning record in 2016?

  • No (51%, 83 Votes)
  • Yes (49%, 80 Votes)

Total Voters: 163

Loading ... Loading ...
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).

178 thoughts to “Poll: Jackson Year 2 Record”

  1. Current total 17 wins
    Healthy Melo full year +5 wins
    Monroe over Jah and Bargs +10
    Corey Joseph over current crew +7
    Top rookie +5
    Better bench production (because they’ll be in a role that fits them) +5
    Knicks type silliness -8 wins.

    Grand total: 41. (a third option needed on the poll)

  2. I think it is basically a question of, “What do you think are the odds that Jackson has a phenomenal offseason?” I think the odds are probably not great, so I guess I’ll go with no. Hopefully Jackson beats the odds.

  3. Uncle Phil said he wants a Big who plays D, I really don’t think Love is in his wish list. I think a lot of people are underestimating Phil’s basketball IQ. He wants to sign 3 FA’s, two of them starters. He also mentioned that DEFENSE is a priority for him. If he can get a player like KAT with our pick, 2 FA’s with a defense mentality and 1 FA who can help Melo in the offensive end, I think we can have a winning record. My wish list is (D. Jordan, Dray Green, M. Carroll, Dragic, Gasol) but a more realistic scenario could be (Monroe, RoLo, Carroll, T. Harris)…. my answer to the Poll is YES….

  4. The Eastern conference is a lot weaker than the Western conference, but it still has quite a number of reasonable teams that one would expect to make the playoffs. Chicago, Cleveland, Washington, Toronto and Atlanta all will probably be over .500 and make the playoffs next year. Boston and Milwaukee are improving and will also probably be over .500. Brooklyn, Indiana and Miami may not be over .500, but are likely to be contending for the last one or two playoff spots. We probably have to be better than all of these last three teams to be over .500 for the year. That seems a tall order to me, so I voted no. I would love to be wrong.

  5. I think a lot of people are underestimating Phil’s basketball IQ.

    Possibly, since I’d have it at 2. (It would have been lower but I have to save room for Fisher).

    That’ll be the next poll. What is Phil Jackson’s NBA GM IQ:
    1
    2
    3

  6. The better question is will Denver swap picks next summer, or just let Toronto have our lottery pick?

  7. If K-bloggers were shareholders in a public company, I could see them all dumping the stock en masse after one year with a new CEO who gets off to a rough start.

    Patience, fellas, patience.

  8. I’m more concerned about post Phil and post Melo than I am the next 3 years. Melo is good. Phil has cap space and a top 5 pick. The cap will go up the following year. We will put together a respectable team, which is all any fan can ask for.

    The real worry is when Phil retires. So far he’s made it clear he will not trade away first round picks and wants to bring back youth and potential in trades. So far that’s been mainly 2nd rounders, D-Leaguers and Euro stashes…and it will take time for those to pay off. But eventually they will.

    The real worry is in 3 years when Melo is near the end and our top pick is the main guy and we’re good but not great and Phil wants to retire. Will Dolan make some stupid trade of our secondary players and future picks to bring in 37 year old Lebron?

    I don’t expect Phil to build a championship team. I just want us winning games and making the playoffs and not being a laughingstock of the NBA. But this new culture he wants to implement, even if he does implement it, will it be strong enough to maintain itself after he leaves?

  9. As time told Jeremy Lin sucked as a player….

    Not true. He’s not a superstar (and I suppose very few thought he could be), but he’s a useful player in the right system.

  10. @Farfa….Maybe I went to far…I know is a touchy subject for lot of guys in this blog…The point I’m trying to make is that for me is even more absurd to say that Phil Jackson has low basketball IQ….

  11. How is a healthy melo worth 5 more wins but corey joseph is worth 7?

    @3, he produced 3.5 this year, and based on his career 3.5+5 is generous.

  12. I don’t think anyone thinks Phil has a low basketball IQ but not all great coaches make great executives. The jury is still out on Phil but so far he deserves a C grade at best. The worry that I have is that Phil has too much confidence in himself and his system, the triangle, and he thinks he can make mediocre players better. The idea that you or your system can improve players is faulty and gets many new executives into trouble. If Phil signs players that have had success and not players that fit his system or “show promise” then he will be okay.

    I actually think he will be fine, not exceptional, but adequate this offseason. He inability to squeeze extra value out of any of his moves so far tells me he probably won’t find hidden value in the market but I don’t think he will make any terrible moves.

    Unless our rookie is really good really fast I would say 35-40 wins is about what I expect next year.

  13. @3, he produced 3.5 this year, and based on his career 3.5+5 is generous.

    How many did Corey Joseph produce

  14. Never, ever tell Knickerblogger regulars to be patient. We were asked to be patient with Houston’s recovery, Curry’s conditioning, Isiah’s batshit-crazy rebuilding plan/trade spree, Grunwald’s post-Lin plan, Carmelo Anthony’s teammate-improving and now Phil Jackson’s nonsensical plot toward relevancy.

    I mean, couldn’t we reasonably ask Derek Fisher to set aside his stupefying ego and turn away for 96 minutes while his team coasted to a near-50% chance at a top 2 pick? That would be asking for patience, and I don’t think it would be unreasonable.

  15. @ 16…. +1 bigger issue with Phil is less the ability to get some players who are clear upgrades over current roster, and more whether the triangle is the right offensive strategy in today’s nail threes and dive to the basket NBA. And is Phil too stubborn to course correct if we start to realize it’s the latter

  16. If K-bloggers were shareholders in a public company, I could see them all dumping the stock en masse after one year with a new CEO who gets off to a rough start.

    If I dumped my fan stock one year into Scott Layden’s tenure, would I have been stupid? Impatient?

  17. Corey Joseph is very similar production wise to Goran Dragic, who is probably good for 8/9 wins over a replacement level point guard. Plus, he’s young enough to learn to shoot a bit better, which would make him a borderline all star. If Melo can stay reasonably healthy and get his 3pt% back to elite territory that’s about the same level of production, assuming Corey Joseph isn’t the beneficiary of some sort of spurs magic and is in better cardio-vascular shape than a certain tooth missing hillbilly.

  18. If I dumped my fan stock one year into Scott Layden’s tenure, would I have been stupid? Impatient?

    Well, if you think Phil Jackson=Scott Layden, then I guess we’re doomed

  19. The good thing about the Knicks is that we gave tons of minutes to guys who are among the worst players in the NBA at their position (namely, Jah, Bargs, fucking Timmy, and Wear). We also got rid of some guys mid season who had be very unproductive. It should, at least in theory, be very easy to improve significantly by replacing those guys with decent/mediocre NBA players. Add in two above average free agents, draft well, and we’re gonna be around .500+ if we get some decent health.

  20. I’m having a hard time thinking of a way to get Joseph out of SA when they have all that cap space. Do you give him a Parson’s deal with less money($7.5-8.5m). Does that even make sense.

  21. Well, if you think Phil Jackson=Scott Layden, then I guess we’re doomed

    From Layden’s wikipedia (which is startlingly long for such an inept GM):

    “On July 23, Knicks shooting guard Allan Houston re-signed as a free agent to a 6-year, $100M guaranteed contract that would keep him in New York until the 2006-07 season, when he would be 35. The signing made Houston, who had never averaged 20-points-per-game in a season at that time, the highest-paid player in franchise history. This transaction put the team “well over $80 million in payroll this season [2001].” Houston retired after the 2004-05 season with chronic knee complications; there was roughly $40,000,000 left on his contract.”

    So Jackson signs an overrated player who’s over 30 to the largest contract in franchise history (and which will also expire when he’s 35) and we’re not allowed to compare him to Layden?

  22. Well, if you think Phil Jackson=Scott Layden, then I guess we’re doomed

    I think they both work(ed) for the same boss.

  23. I’m having a hard time thinking of a way to get Joseph out of SA when they have all that cap space. Do you give him a Parson’s deal with less money($7.5-8.5m)

    The Spurs are (probably) going to want to bring Danny Green back, max-Kawhi, and then resign or replace Manu and Timmy. They have 17 or 18 million tied up in the PG position in Mills & Parker. Are they going to want to put another 8 million into the position? I’d at least find out.

  24. Carmelo was 29 when he signed this deal which lasts five years until he is 34. Houston’s six year deal lasted until he was 36. And I think Carmelo is a more versatile player (who has scored more than 2o points a game too.). But the comparison is still fair enough to be worrying.

    But I don’t think d-mar was saying the comparison was unfair. He was just making the statement that if Jackson works out the way Layden did, that’s really bad. I can’t argue with that.

  25. How many did Corey Joseph produce

    According to WOW 6.3 in 18 minutes. Not bad. If he can keep it for 36 he’d be at the 12-13 range and the current Knicks crew has maybe 7-8 together.

  26. –According to WOW 6.3 in 18 minutes. Not bad. If he can keep it for 36 he’d be at the 12-13 range and the current Knicks crew has maybe 7-8 together.

    where do you go for this? I tried to google it but no luck.

  27. You can get wins produced numbers at boxscoregeeks.com

    I realize that’s not everyone’s cup of tea, which is why I think it’s helpful to imagine Corey Joseph as Goran Dragic, only 23 years old.

  28. “The signing made Houston, who had never averaged 20-points-per-game in a season”…. Until we see what Uncle Phil does this offseason, I think we can’t compare him to Layden…. Did Houston ever won a scoring tittle? Was he ever in the top 5 in the MVP voting in a season? The hate on Melo on this board is REAL…lol…

  29. How come every comment section on this blog turns into someone stumping for the Knicks to sign Corey Joseph? Not well versed enough on his per minute stats to say that he would be a bad signing at $8 million but I’m just curious with all the available FAs how people have become so fixated on Joseph.

    Also curious about how people would rank preference for signing these FA at the starting mini-max of $15,768,500….Greg Monroe, Khris Middleton, or Reggie Jackson? Shockingly I believe that all three will get that deal from someone.

  30. Did Houston ever won a scoring tittle? Was he ever in the top 5 in the MVP voting in a season?

    Well, if Jackson is looking at ppg and MVP voting to evaluate players, he may as well be Scott Layden.

  31. didn’t Wages of Wins have Landry Fields in the top 10 for win shares his last year as a Knick?

    talk about flawed methodology….

  32. @Donnie Walsh we can mention All Star games selections, All NBA 2nd and 3rd Team selections…etc…. I was just trying to establish the point that Houston and Melo are not in the same category as NBA players. Nothing more, nothing less…

  33. Greg Monroe, Khris Middleton, or Reggie Jackson

    Monroe > Middleton > > > > > > > > > > > > > > >Jackson.

  34. didn’t Wages of Wins have Landry Fields in the top 10 for win shares his last year as a Knick?

    talk about flawed methodology….

    Mmm, feels like 2011 all over again.

    http://www.boxscoregeeks.com/players/220-landry-fields

    No. I don’t even want to start with this conversation, but “I don’t agree with the results, so the methodology must be flawed” is some functionally-illiterate, argument-challenged, lowest-common-denominator, average-American-voter, what-me-worry sort of nonsense garbage methodology if I’ve ever seen it. And I’ve seen it a lot on this board.

  35. we can mention All Star games…

    You forgot to mention Slam Dunk Competitions! (Allan Houston came in 4th. Carmelo?)

    I was just trying to establish the point that Houston and Melo are not in the same category as NBA players.

    I think they are closer than you think, not just in the fact that the Knicks outbid nobody to give them the max, but in their actual basketball production too.

  36. Jowles, i understand your frustration. I think where others have issues with these catch all stats , especially WP is the outrageous outliers that it produces. Its hard for people to just take the stat as gospel when it gives results like that. I have come around to advanced stats but i try to use them as a check for what my eyes and the box score are saying. Thats why i had an issue with the Melo/Harden TS% comparison, it just doesnt jive with the box score for me but what do i know?

  37. Allan would throw down a ludicrous dunk and then go the next 4 games without ever driving to the rim. Used to drive me fucking nuts. I believe he was the player that got me thinking pointz might not be the only way to evaluate a players full worth.

  38. –I think they are closer than you think

    i dont think so, Houston was meek as hell and needed Spreewell to be the alpha on those teams. Houston was much more of a role player in every sense.

  39. We talk about CoJo a lot because he’s a promising young player who’s a realistic target for the Knicks and projects to be a great value at the 6-8 million range since he’s basically Dragic but better on d.

    DRed you’d take Monroe over Middleton? Interesting.

    I’d marry Middleton fuck Monroe and kill reggie Jackson

  40. I think Melo is a better player than Houston was, but they share a frustrating inability or unwillingness (or whatever) to translate their basketball skill into effective play.

  41. Jowles, i understand your frustration. I think where others have issues with these catch all stats , especially WP is the outrageous outliers that it produces. Its hard for people to just take the stat as gospel when it gives results like that. I have come around to advanced stats but i try to use them as a check for what my eyes and the box score are saying. Thats why i had an issue with the Melo/Harden TS% comparison, it just doesnt jive with the box score for me but what do i know?

    Here are the top players in WP (total, not rate) this season:

    Paul
    Jordan, DeAndre
    Curry
    Harden
    Davis
    Chandler
    Butler
    Gobert
    Leonard
    Gasol, Pau
    Korver
    Duncan
    James, LeBron
    Green
    Bledsoe
    Lawson
    Lillard
    Green
    Wall
    Westbrook
    Drummond
    Batum
    Carroll
    Irving

    That looks like a list of players that pretty much anyone on this board would love to have on the Knicks. I’m not saying that your subjective basketball opinion matters, but I don’t think it’s too far off of WP’s conclusions. The reality that a player like Carmelo might not make as many shots as an elite player? Or that he might actually turn the ball over? That’s hard to swallow for many of y’all.

    This is my last post on this because I’m getting all agitated again. WP isn’t perfect. Carmelo is not as good as his contract or all-star appearances or scoring title might suggest.

  42. Anybody like Donald Sloan from the Pacers? His salary was just under 1mil this season and can be had at a fraction of the cost of some of the common PG targets out there. His shooting pct doesnt look very sexy but could be a solid backup PG for us.

  43. “I think where others have issues with these catch all stats , especially WP is the outrageous outliers that it produces.”
    I could be wrong, but I think that Jowles has, on occasion, acknowledged that, once in a great while, even his favorite catch-all stat produces an outrageous outlier or two. What is fascinating is that he appears to have never once even considered the possibility that perhaps Tyson Chandler is one of those outrageous outliers…
    By the way, the funniest thing about the Layden wikipedia page is that 15 years ago, the Knicks had a payroll of over $80 million. Hardly anyone today has a payroll that high.

  44. What’s an outrageous outlier? That Tyson Chandler rebounds a shitton of misses and almost always scores when he makes an attempt to? As I’ve said before, you have to do a lot of leaping to get from “Chandler is an efficient scorer” to “Chandler’s low volume shooting forces bad shot attempts from other players on his own team” or “Chandler’s paint presence results in a lot of high-quality shot attempts” to “Chandler’s paint presence makes it harder for his teammates to score the basketball on such a quantity of available possessions that his high efficiency is not indicative of high productivity.” That’s a lot of leaping.

    No, you’re right, Capt. Luke. LeBron has never missed a shot or turned the ball over, and when he does (but remember that he doesn’t), it’s less bad than when other players do.

  45. I don’t see how ranking Chandler et al over LeBron is beyond the intuitive pale given that Chandler is having his best season ever and LeBron is having, by his standards, a below average one (the first third of his season was way below his standards before he rounded into form).

    I think LeBron could be better valuated by WP but just because it doesn’t fit your intuitions doesn’t really mean much–Dennis Rodman was better than Michael Jordan during some seasons, regardless of your intuitions to the contrary. So it could be for Chandler or Gobert.

    Furthermore WP is likely undervaluing Chandler and Gobert’s contributions because it kind of punts on defense, which is their major strength.

  46. Generally speaking WP only produces the Bismack Biyombo/Landry Fields esque outliers when you look at it on a rate basis. Though, to be fair, Biyombos way better than people give him credit for and so are probably most of the people who are overrated by WP.

  47. so is saying Joseph had 6.3 Wins Produced really the equivalent of 6.3 Wins Above Replacement? With the league depth at PG right now I don’t see how that could be possible.

  48. —-I think LeBron could be better valuated by WP but just because it doesn’t fit your intuitions doesn’t really mean much–Dennis Rodman was better than Michael Jordan during some seasons, regardless of your intuitions to the contrary. So it could be for Chandler or Gobert.

    Great points. Which is why i am more accepting than in my youth.

  49. OKC fired Brooks. Now, if they learned something, they will hire David Fizdale or Shaka Smart, or even Jim Boylen. But why not Derek Fisher? He was one of the boys there! Come on Presti, do us a solid!

  50. “What’s an outrageous outlier? That Tyson Chandler rebounds a shitton of misses and almost always scores when he makes an attempt to? As I’ve said before, you have to do a lot of leaping to get from “Chandler is an efficient scorer” to “Chandler’s low volume shooting forces bad shot attempts from other players on his own team” or “Chandler’s paint presence results in a lot of high-quality shot attempts” to “Chandler’s paint presence makes it harder for his teammates to score the basketball on such a quantity of available possessions that his high efficiency is not indicative of high productivity.” That’s a lot of leaping.”
    QED…

  51. I was skeptical at first about WP. But then I realized something: that a very efficient intern is way more valuable to a company than a talented manager who gets easily bored and wastes hours upon hours playing Candy Crush instead of looking at that company’s problems.

  52. The thing is that every catch-all advanced stat produces strange outliers–it’s just by virtue of how you choose to weight certain inputs: PER is weighted in favor of volume shooting despite middling or below average efficiency; WP rewards low usage high efficiency scoring and rebounding; WS confuses team defense for individual defense and rewards sub par defenders on good defensive teams. Just because the WP outliers seem gaudier doesn’t mean that it’s necessarily more flawed (though it could be). The presence of outliers is not a compelling reason to reject any particular catch-all metric and certainly isn’t an in principle reason to reject all of them. It’s more of an invitation to examine the metric and see if the presence of an “outlier” is either reason to adjust the methodology of the calculation or if it’s a reason for us to reassess a player we saw previously as not very good.

  53. To be honest, Chandler doesn’t swat all that many balls. But he perfected the concept of “team defense dominating center”, then perfected even more by Marc Gasol.

  54. Whatever you do, do NOT blame Chandler for the fact that the Mavs have given up 229 points in 2 playoff games.

    His teammates suck on defense, that’s all there is to it.

  55. @62 right. Thats why these stats have to be used additional frame of reference not the rule.

  56. Outregeous is to give more credit to a very limited player like Chandler versus a great player like LeBron. Chandler is great at the VERY FEW things he does, like scoring one foot from the basket and staying in the paint swatting balls in any direction.

    I don’t think having a wider range of “skills” is strongly correlated with being a better basketball player. Melo has a lot of “skills” that he fails to leverage into great play; Tyreke Evans has a lot of “skills”. Chandler is bonkers at all the “limited” things he does, like rebound, defend, and shoot. To me being elite in those categories hardly counts as “limited” unless you’re saying that because he can’t put the ball on the floor or work in the post he’s somehow less valuable than people who can but that doesn’t follow logically or empirically.

  57. Wouldn’t you know? WP says that Lou Amundson produces value equal to an average center, while both Jah and Bargs produce negative wins. Doesn’t that sound right?

  58. @66 uh yes. They were remarkably dreadful yesterday. They made Josh Smith look like Magic Johnson.

  59. I’m pretty sure the guy at the door at Home Depot whose job is to make sure you produce a paid receipt upon leaving the store makes far fewer mistakes than the Store Manager. Which one is more efficient? Who is more valuable to the organization? Which one is easier to replace? It’s all about skewing the numbers.

    They’re equally easy to replace because all retail management requires is a pulse and a half-functioning brain stem. Hierarchical management is dumb even though everyone does it–worker self-management or a horizontal corporate structure like at Valve is generally way better.

  60. I don’t think having a wider range of “skills” is strongly correlated with being a better basketball player. Melo has a lot of “skills” that he fails to leverage into great play; Tyreke Evans has a lot of “skills”. Chandler is bonkers at all the “limited” things he does, like rebound, defend, and (Dunk/Layup)

    –Fixed that for you.

  61. Also, Tim Hardaway Jr produces negative wins while Pablocito produces all the wins in the world.

  62. Exactly. Statistics are extremely valuable when used to further double-check the eye test (the mind is the most complex statistical analysis tool known to man, but that’s a simple fact that the math warriors will not acknoweledge).

    That’s actually patently untrue because human psychology is subject to all kinds of biases and misconceptions that make us radically bad at cost-benefit analysis or perceiving the overarching statistical implications of games and seasons. The human brain also lacks the computational power of a statistical model unless when you watch basketball you’re calculating everyone’s statistics and doing a real time regression to the mean in your head but nobody does that when they watch basketball. The brain is much better than a computer at most things, but stats is not one of them.

  63. LeBron but we’re talking about who’s having the better season because that’s what stats are capturing not “potential” or historical performance. Just because I would put LeBron on a contract before Chandler doesn’t mean that Chandler isn’t having the better season or has been the better player, this season (since again that is what the stats are measuring).

  64. The human brain is a far, far, far more sophisticated calculating machine and produces far, far, far more accurate results than any statistical method in existence.

    You need look no further than where the idea that statistics could be more valuable than the eye-test originated to determine that this statement is totally false: in baseball, the eye-test cannot tell the difference between players who hit .220 and .280, or get on base .290 or .380, but the difference is enormous.

    Human beings experience emotion. A player who makes a crazy, off-balance 3 with time running out will be perceived as better than a player who makes a high percentage of un-dramatic 2 point shots while being fouled occasionally. Computers do not experience emotion, and do not give style points that are unavoidable in the eye-test.

    That said, no one has said that any existing statistical model is perfect. But they are informative and illuminating, even if flawed.

  65. store manager to door guy is a pretty flawed analogy, maybe better would be ::Melo:Chandler —> ::Melo:that guy who pours Melo’s gatorade without spilling. Fundamentally, maximizing the number of your team’s possessions and the number of points scored per possession are what wins you basketball games. Skills like being able to “create your own shot” are not terribly useful if said shot is a 41% 20-footer. Just because the rest of your team is incapable of creating a better shot than that on a routine basis is not indicative that it’s a useful skill, just that your team sucks.

    By the way, I’m not a huge fan of WP and believe it to be very flawed, but the underlying premises (which no statistical measure can model very well) are sound.

  66. You can also look at the results of something like Wins Produced (which was invented, I should add, by a person using their brain) and then use your brain to analyse the results.

  67. The model comes from someone thinking it up, but the actual computations are done by a computer, because a human can’t do them fast enough, or consistently enough. Those two things are conceptually separate.

    The human brain is a far, far, far more sophisticated calculating machine and produces far, far, far more accurate results than any statistical method in existence. In fact, the only field in which computers outdo the human brain (as of today, may change in the future, I have no idea) is brute, simple calculation. Far more complex tasks, determining that the two guys at the corner must be up to something is a task totally beyond the computer’s number-crunching abilities.

    Powerful chess AI is decades old and is better at computing possible moves than the best grandmasters in the world and that was with a brute force algorithm that has since been refined into something that embarrasses the prior algorithm. If the brain was better at computing than a machine, then why can’t it calculate more chess moves in advance than a computer? It’s because the human brain is categorically worse at computation.

    Your example of “determining that someone is up to something” is not a computational issue, it’s an issue of inductive problem solving that carries with it a lot of non-mathematical freight accrued over decades of experience which is why the brain is better at it than a computer would be.

    The brains computational power is mediocre, but it more than makes up for that with inductive problem solving and conceptual analysis that computers can’t do. But that still doesn’t mean the eye test is accurate or useful in a vacuum because the eye test is an attempt at computation and mathematical quantification from observed data, and computers will, always and forever, be better at tracking and quantifying and computing information off of that. Our job lies in the analysis of what the computer spits out.

  68. You know, Tyson Chandler has produced 16 wins this season, but Dallas has won just 1 game more than last season.

    JR Smith went from 0 WP48 to .170 WP48 changing teams this season.
    Josh Smith went from -0.2 to .140

    wild fluctuations for a metric, huh?

    For all his flaws, PER does not fluctuate so much.

  69. Basically all I’m saying is that the brain is not a computer and shouldn’t be treated like one and that conclusion cuts both ways–it’s worse at computation than computers but better at lots of other things. Qualitative value judgments are important in basketball and shouldn’t be swept aside but when you champion the “eye test” as the end all metric for player performance because of some dumb notion that the brain is better at telling the difference between a player with 10% TRB and one with 18% TRB than a computer is missing how and why the eye test is valuable and reveals a fundamental misunderstanding of computers, stats, theory-making, and how the human brain works.

    The eye test is not computationally valuable and it’s not rigorous enough to overcome human psychological biases and so isn’t that useful in forming foundational intuitions about players vis a vis other players (e.g. Chandler is way better than Carmelo though that flies in the face of eye test intuitions)–it’s use is found in interpretation and explanation. It’s not super great at evaluation beyond the general sense of a person’s “game” that is important to consider when looking at team fit but doesn’t have much use beyond that.

  70. JR Smith went from 0 WP48 to .170 WP48 changing teams this season.
    Josh Smith went from -0.2 to .140

    wild fluctuations for a metric, huh?

    For all his flaws, PER does not fluctuate so much.

    uhhh JRs PER rose by 3 which is huge fluctuation for PER and his WS/48 rose from -.008 to .124. All of his advanced stats fluctuated not because they’re oversensitive but because he’s way better in Cleveland than he was here. Same with Josh Smith, though his fluctuations are less dramatic (+1 PER, +.050 or so WS/48).

  71. “You know, Tyson Chandler has produced 16 wins this season, but Dallas has won just 1 game more than last season.
    You’re not supposed to point that out here, nor are you supposed to point out the fact that he has led the league in offensive rating for something like 4 of the last 5 years, yet the Mavericks were worse on offense this year than last. Here’s the thing — some people on this site rave about what an offensive machine he is. You know how many times in the last 4 years he has scored 20 points or more? 13 times. In those four years, there were loads of games in which his team was struggling to score and really could have used an efficient offensive juggernaut to step up, put the team on his back, and carry them. Did he do it? No, not once. Why? Either because he was selfishly protecting his efficiency stats so he stubbornly refused to shoot more, or he is totally incapable of doing so. It was probably the latter since, except for when he obviously gave up on his team last year, he doesn’t strike me as a selfish player.

  72. Uh contemporary chess AI isn’t exclusively brute force and utilizes a lot of selection algorithms and chess ai hasn’t gotten better because of faster processors but because of more efficient algorithms so it’s not ” a very simple game of brute-force calculation”.

    The former problem is, again, not computational and so it’s not something we would fault a computer for not being good at it. The latter problem is part computational, part interpretive and we can’t adequately answer it without both, but, as I’ve said before, stats are argumentatively prior to eye test evaluations.

    No one is saying that LeBron isn’t better at basketball than Chandler, we’re saying that Chandler has had a better season. This holds true when you look at the stats and really illustrates the issues with “intuitions” and “common sense” and the eye test that we’re clearly never going to agree on.

  73. “Skills like being able to “create your own shot” are not terribly useful if said shot is a 41% 20-footer.”
    It’s a useful skill if the “efficient”guy on your team absolutely, positively won’t/can’t take a shot if it isn’t a dunk so you are stuck taking the 41% shot when the shot clock is expiring.

  74. Eye-test + stats > stats.

    No one is saying to get rid of the eye test it’s just a recognition of the fact that the eye test has limited use in player evaluation and does a fundamentally different job than statistical analysis–they should coexist because they don’t even do the same argumentative work and don’t have to be in tension with one another. It’s when people think the eye test has a privileged position vis a vis stats in player evaluation that a tension arises because there’s a mistake made in the assumption that these two things do the same kind of argumentative work.

  75. “No one is saying that LeBron isn’t better at basketball than Chandler, we’re saying that Chandler has had a better season.”
    Do you really think that Chandler “produced more wins” for his team this year than LeBron did? Seriously? One would think that, if that were the case, his presence (and brilliance) would have led them to a better record — especially since the guy he replaced was Dalembert.

  76. Well it doesn’t help that Dallas has a pretty bad team otherwise and Dirk has fallen off massively–basketball is a team game, guys.

    Also it’s more because LeBron has been below average this season more than because Chandler has been transcendent. Chandler is having his best statistical season on record, and LeBron is having one of his worst–it’s not beyond the pale to think that he could be better this season even though LeBron is usually in another class.

  77. If we fielded 5 Chandlers it would mean that we had developed the ability to clone humans, single-handedly pushing the technology of all humanity forward by decades. I think that’s a little more important than how many points we would score in the next game. Try to see the big picture guys.

  78. There is no weaker argument against WP (or any other metric) than merely pointing out what one considers outliers. If you start with the premise that WP is wrong about Fields, Chandler, or anyone else, then yeah WP will look wrong to you in many cases. If you actually examine the methodology and see for yourself where those guys derive their value as far as WP is concerned, then you can make an argument against it (if you still disagree).

  79. lol who is advocating for 5 chandlers, or that WP is an absolute measure of a player’s worth in isolation? the stuffing in the scarecrow is a bit ragged.

  80. Jowles reading this thread=Excedrin headache.

    I will again make this argument re: efficiency:

    Is a 7-footer that only goes after rebounds he is nearly sure to get, but only averages 5 rebounds per 36, celebrated for being an “efficient” rebounder? Imagine saying Steve Novak did a heck of a job on the rebounds he got, and a great job letting the better (but less efficient) rebounders get theirs. So what that he averaged less than 4 rebounds/36 at 6’10″…he’s a great rebounder!

    Cherry-picking is cherry-picking. Chandler is the ultimate cherry-picker. (among starting players, has the lowest usage rate in the NBA!!!). He is just as flawed as a 7-footer who shoots at a TS% of .650 at a 30% usage but averages 1 rebound a game.

  81. The major beef I have with WP’s valuation of Chandler et. al. is that it assumes (in my view, wrongly) that the statistical factors that are most highly correlated with wins actually cause (or produce) the wins. Higher scoring efficiency correlates with wins better than lower scoring efficiency, regardless of usage. The key factor that makes Chandler’s WP so high is that he is a great offensive rebounder, and he doesn’t lower the impact of his offensive rebounding by missing shots he is unlikely to make. He lets others do the missing for him.

    PER is sort of the mirror image…it over-values volume scoring to the point where big scorers are likely to get high PERs, regardless of efficiency. But it (in my opinion, correctly) penalizes the extreme low-usage player, regardless of efficiency.

  82. Two points about computers:
    1. The brain is 100% a computer: a complex, biological computer consisting of hierarchical system of pattern recognizers.
    2. Computers do tons of things better than humans. These things included brute force computations, but also things like predicting what movies someone will like and—some time in the near future—driving cars on busy roads. Computers have two major advantages: computational power and memory storage/better data. With these two advantages, it is only a matter of time until computers do nearly everything better than we do. We might even see people build an artificial human brain (down to the molecular level) in our lifetimes.

    I think the eye test is best at discovering effects/interactions that the computational models don’t handle, and using that knowledge to improve our models or at least make us aware of those deficiencies. Right now, I think one of the biggest deficiencies is determining how different lineup will perform, both in general and versus other specific lineups. Issues like: how can Dallas field a lineup that both spaces the floor on offense at matches up with Harden and Howard on the other end? Where can they hide Dirk? Who should they play at PG? I think these are the most interesting questions in basketball and part of what makes basketball so exciting: the coaches and the lineup selections have a tremendous impact on the outcome of games, especially playoff games.

  83. If we fielded 5 Chandlers it would mean that we had developed the ability to clone humans, single-handedly pushing the technology of all humanity forward by decades. I think that’s a little more important than how many points we would score in the next game. Try to see the big picture guys.

    haha, I laughed out loud, thank you

  84. I’m not arguing against WP because there are some outliers; I am merely arguing that Tyson IS one of WP’s most glaring examples of an outlier. He’s a good player, top 10 in the league at his position, but clearly not top 10 overall.

  85. He’s a good player, top 10 in the league at his position, but clearly not top 10 overall.

    Based on what?

  86. Also, advanced basketball stats have nothing to do with computers. This is getting embarrassing

  87. I remember that bit in Thinking Fast and Slow about how difficult it is to manipulate multiple numbers in your mind at once. And then dudes on this board being like (and I paraphrase), “Yeah, the brain can tell you that Tyson Chandler is not a top 10 player, but Carmelo? Obviously.” It’s like c’mon dudes let’s get fucking real with it.

  88. The major beef I have with WP’s valuation of Chandler et. al. is that it assumes (in my view, wrongly) that the statistical factors that are most highly correlated with wins actually cause (or produce) the wins.

    Yeah, which sounds like a pretty fair assumption to me. I understand correlation isn’t causation obviously, but when 30 years of data told us what factors were most common in won basketball games, I’m inclined to think that’s at the very least a good start.

  89. Face,
    I’m not arguing that WP is dumb or meaningless. I just think that it overvalues and undervalues some things. Clearly, Chandler is a difference-maker who has some elite skills and plays within himself. But WP ignores the holes in his game and blames other players who have to make up for the things he can’t do. Again, HE HAS LOWEST USAGE RATE OF ANY STARTING PLAYER IN THE LEAGUE! Yet, WP ranks him as an all-time great offensive force.

    Again, give me a thoughtful response as to why WP doesn’t similarly reward rebound cherry-pickers for not going after rebounds they are not as likely to get?

  90. My argument remains (for years now) that it is pointless to compare the value of players with dramatically different roles without considering the needs of the team. In WoW’s world, the 90’s Knicks would have won more games with Chandler at C rather than Ewing.

  91. Again, give me a thoughtful response as to why WP doesn’t similarly reward rebound cherry-pickers for not going after rebounds they are not as likely to get?

    You don’t use a possession when you try and fail to get a rebound, you do when you try and fail to make a shot.

  92. One place where I have to give WP is in finding value in a salary cap driven league. It is easy to be blinded by flashy scoring or volume scoring, and to grossly overpay for it. That’s why I agree with Dean Oliver that (to paraphrase) you can’t use any stat in a vacuum. Team needs, playing style, and interaction effects all have to be factored in, and the best approach is to use is to combine statistical analysis with a trained eye.

    I also (somewhat begrudgingly) agree that the trained eye alone will make more mistakes than the statistical analysis alone. But the larger point is that both will make plenty of mistakes, and in the high “impact per decision” world of NBA basketball, minimizing mistakes and maximizing valuation is key (as the Spurs have demonstrated…maybe Anderson was a mistake in the 2014 draft, but they are so far ahead of the curve, so what?)

  93. But WP ignores the holes in his game and blames other players who have to make up for the things he can’t do

    Who is “making up for the things he can’t do?” Do players see an increase in usage/FGAs when they play with Chandler historically (that’s not rhetorical)?

  94. It would seem to me that 100% of possessions have to be “used.” If things were perfectly balanced, all 5 players would have a usage of 20%. If one player is down around 12%, someone has to make up for the other 8%.

  95. In WoW’s world, If (to use round numbers) Melo uses 30% and Chandler 10%, Melo is hurting the team, because he is using the 10% variance from average at a much lower efficiency. In a perfect WoW world, Chandler should use 100% of possessions!

  96. It would seem to me that 100% of possessions have to be “used.” If things were perfectly balanced, all 5 players would have a usage of 20%. If one player is down around 12%, someone has to make up for the other 8%.

    But have players with otherwise lower usages been “forced” to up their usage over the course of a full season due to playing with Chandler? I can’t find an example of such a thing happening.

  97. Here’s a way to think about it, Z-man. The average NBA starting center is going to score a few more points per game than Tyson. So Tyson’s team is going to be down 3 or so points compared to an average NBA offense with an average center. But because of Tyson’s efficiency and his offensive rebounding, they’re going to have about 7 or so possessions to make up those 3 points. NBA teams average about a point per possession, so while in some sense there is an increase in the burden on the rest of the guys on the Tyson team (about a shot a game), it is very likely that teams offense will do better than one with an average center.

  98. If we fielded 5 Chandlers…

    I’d take a lineup of Tyson Chandler, Wilson Chandler, Raymond Chandler, Chandler Parsons, and Chandler Bing over the 2014 Knicks anyday.

  99. @122 so you agree that Ewing was closer to being an “average” center than Chandler?

    My beef isn’t with anyone who says that Chandler is “better than average” with which I wholeheartedly agree. It is with those who (based on WP) consider him an all-time great.

    Again, Chandler has some elite skills, but has a glaring flaw that is not picked up by WP.

  100. It’s pretty amazing how Knickerblogger experts predict our win totals for next year while our roster is still under construction.
    So why lets also predict who will be NBA 2020 champions.

  101. Again, Chandler has some elite skills, but has a glaring flaw that is not picked up by WP.

    The point of being good on offense is to help your team score more than the other team, right? Say the other team scores 100 points. If Tyson scores 14 points on 6 FGA, it is easier for his team to score over 100 points than it would be if Tyson had scored 18 points on 14 FGAs.

  102. Holy cow, Brooklyn down 1 with 28 seconds left and Millsap at the line. Either Brooklyn is playing above their station, or Atlanta is playing well below theirs.

    EDIT: Millsap goes 1-for-2. Brooklyn with the ball, down 2 and 27 seconds left.

  103. Deron Williams misses a wide open 16-footer that would have tied the game. Korver sinks two FT’s and seals the win.

  104. I tried to find the Nets game, but it was blacked out on NBA tv and trying to figure out what station was showing it reminded me that I don’t give a shit about the Nets.

  105. “If Tyson scores 14 points on 6 FGA, it is easier for his team to score over 100 points than it would be if Tyson had scored 18 points on 14 FGAs.”

    Well, that would assume that he scored 4 points on an extra 8 FG attempts. The question is, What if everyone on the team set the efficiency bar that high and the usage bar that low?

    In your example, if the guy who scored those 4 points on, say 4 shots, he would be seen as relatively inefficient, even though it might have taken Chandler, say, 6 shots to get an additional 4 points. When his teammates have a relatively inefficient night, Chandler gets none of the blame. He protects his efficiency by simply not shooting.

  106. Wins Produced breaks down at the extremes of usage.

    There’s no doubt that shot selection is often an issue with some of the higher usage lower efficiency scorers (like Melo). Their rating should be penalized for that poor shot selection to the extent that it’s a self inflicted problem. But often, players like that also have a greater role in the offense because they are more versatile scorers than other players on the team. So on some occasions it makes more sense for them to take some of the marginal 2 pointers that defenses force upon a team than some other players that are more limited. They should not get punished for their role.

    On the flip side, there are some extremely limited offensive players who can be fairly efficient as long as they just score at the rim, on breakaways, at the FT line etc… Those players deserve SOME credit for being smart enough to avoid shots they would not be efficient at, but they don’t contribute as much to the offense and Wins Produced suggests. All else being equal, Wins Produced will tell you that some guy that is terrible on offense, but make 3 dunks on 4 shots is more valuable on offense than a player with versatile scoring ability, average usage, and mildly above average efficiency.

    There is some relationship between usage, efficiency, shot selection, role on offense etc… that Wins Produced just doesn’t have quite right. That’s partly why we see some players that rate really well on Wins Produced that barely even play (Fields etc..). It’s not that all these coaches are stupid. These players are extremely limited on offense and Wins Produced is not rating them properly.

    As long as you are aware of this, I think Wins Produced is quite useful other that at those extremes. In my own thinking, I just make mental adjustments to those players.

  107. The 76s, by far the worst offense in the NBA last year, scored .93 points per possession. Every team in the NBA will average more than 4 points on 8 possessions.

  108. If I told you that Z-Bo and Gasol were a combined 5-22 from the field how many of you would have guessed that Memphis would be up double digits?

  109. @132 well said, totally agree. Same is sort of true (different flaws) for PER and WS.

    @133 but what does that prove relative to WP? If the shooters that fill the void left by Tyson’s low usage with 1.o5 PPP, they still get hammered by WP while Chandler gets all the credit.

  110. Wins Produced breaks down at the extremes of usage.

    There’s no doubt that shot selection is often an issue with some of the higher usage lower efficiency scorers (like Meloelse being equal, Wins Produced will tell you that some guy that is terrible on offense, but make 3 dunks on 4 shots is more valuable on offense than a player with versatile scoring ability, average usage, and mildly above average efficiency.

    There is some relationship between usage, efficiency, shot selection, role on offense etc… that Wins Produced just doesn’t have quite right. That’s partly why we see some players that rate really well on Wins Produced that barely even play (Fields etc..). It’s not that all these coaches are stupid. These players are extremely limited on offense and Wins Produced is not rating them properly.

    As long as you are aware of this, I think Wins Produced is quite useful other that at those extremes. In my own thinking, I just make mental adjustments to those players.

    Wins Produced is a relatively simple linear weights metric. To oversimply: the idea is that the metric already does account for the importance of FGA distribution among its players. And the conclusion is that FGA distribution does not matter as much as eFG efficiency does.

    There is no correlation between a team (1) having a player who takes very few shots relative to his teammates and (2) an overall drop in efficiency among his teammates. If there were, the metric would account for that.

  111. “the conclusion is that FGA distribution does not matter as much as eFG efficiency does.”

    Not sure what you are trying to say here…

  112. OK, if Chandler is so awesome, how come the Knicks’ record when he was hurt was no worse (and sometimes much better) than when he played, yet their record when Melo was hurt was much worse than when he played? And how come, when the Mavs replaced a stiff like Dalembert with an all-time great, their record didn’t improve one iota? One would think that a top 10 player would have a little more of a positive influence on his team’s record than that. Unless, of course, he is the outlier who is over-rated by the great infallible all-inclusive stat. Could that possibly be the case?

  113. And how come, when the Mavs replaced a stiff like Dalembert with an all-time great, their record didn’t improve one iota?

    Out of the top 10 Mavs in minutes played last season only 3 are still on the team. It’s basically Dirk, Monta and a whole new team.

  114. So long as Hibbert’s, er, Howard’s valet (no more misogyny) doesn’t shoot, he is immune from blame.

  115. but what does that prove relative to WP? If the shooters that fill the void left by Tyson’s low usage with 1.o5 PPP, they still get hammered by WP while Chandler gets all the credit.

    First, Tyson’s low usage doesn’t create much of a void. Yes, his teammates would have to increase their usage a bit, but not in a way we would expect to have a negative effect on their efficiency. Second, 1.05 PPP is pretty good.

    What Tyson does on offense is help his team score more points. That’s good, right?

  116. I am going to be utterly useless at work tomorrow, but it’s worth it. This is a great game.

  117. Not sure what you are trying to say here…

    I’m saying that if a player’s low FGA rate impacted team wins it would be reflected in a linear weights model. If a player has, say, a low OBP, a linear weights model would determine that he failed to do certain events recorded in the box score that correlate with wins (BB, 1B, HR, etc.). The point is that WP suggests that FGA is not particularly relevant to a player’s value, or it would be indicated by the model as predictive of success or failure.

    PER is a stat that “works” to people because it reinforces the concept that players who bear a shooting volume load are doing god’s work for their team. PER poorly reflects actual wins (even in the past, not even predictively), so it really doesn’t confirm or deny the importance of volume scoring.

  118. Longtime reader here, very rare do I post. I’d just like to ask something that’s confused me regarding advanced stats as applied to basketball. Hopefully the more enlightened among us can help me out!

    As I understand it, the movement towards advanced stats or analytics has its roots in baseball. My take on this is that basketball is a much more complex game than baseball, in the sense that many more factors can affect a basketball player’s performance rather than a baseball player.

    System, chemistry, teammates — all these can contribute to overstating/understating a player’s worth. If a player, for example, who only shot open corner threes because he was teammates with a dominant scorer who demanded double teams, was moved to another team, would he perform up to the level the stats say he should perform?

    I guess what I’m asking is this — how do the advanced metrics existing today account for these differences in situations per player?

  119. Man. If Charles Barkley thought Daryl Morey was a nerd, he would have field day with today’s comments ha ha.

    I guess what I’m asking is this — how do the advanced metrics existing today account for these differences in situations per player?

    I am no stats guru, so I might leave it to someone else to give you a better answer, but I imagine coupling advanced stats with old fashioned eye-test scouting would be the answer to your question.

  120. Fun game, but I dunno, it was also pretty darn sloppy at times on both ends.

    Also, I love me some Chris Paul, but dude has to chill it out with the rage issues.

  121. Great line by Bill Simmons about the game, “This is the most exciting atrociously played 4th quarter I’ve ever seen.” That sounds about right.

    And yes, it is insane that Tim Duncan is older than Brad Stevens.

  122. System, chemistry, teammates — all these can contribute to overstating/understating a player’s worth. If a player, for example, who only shot open corner threes because he was teammates with a dominant scorer who demanded double teams, was moved to another team, would he perform up to the level the stats say he should perform?

    You should use stats to describe, not prescribe. That’s my attitude. I don’t care if Tyson Chandler were a bad basketball player, if the game were to revolve around taking left-handed, blind-folded hook shots. He is a tremendously productive basketball player, and it’s up to coaches to replicate whatever he’s been doing for the last five years and stick fuckin’ to it.

  123. The TNT announcers seemed to believe that the league will try to get rid of Hack-a-Shaq for next season. I don’t see how they will do that. They didn’t do it when Shaq actually played, so why start now?

  124. You should use stats to describe, not prescribe.

    Even if you think basketball can be described by a linear weights model, all correlations that WP finds are at the team level. Then it PRESCRIBES how to distribute the responsibility for the wins.

    A team can only produce so many assisted baskets. Some of the WP superstars are shoot only assisted baskets. When the team is unable to find an assist for Chandler, then the “average player” is expecte to score at the “average efficiency”, because Chandler is much worse than the “average player” at shooting unassisted (33%, when he cherry-picks to shoot unassisted, that is)

    Well, this “average efficiency” of the “average player” is what it is, because every team usually has 2 or 3 chuckers on the field, eating 70% of the possesions, and not letting players like Chandler shoot 33%. But somehow, they don’t get credit for it, because if you compund their efficiency with the efficiency of Chandler’s assisted baskets, they are shooting less than the average!

    So, somehow, WP has decided that it has to penalize players that are better than Chandler, and reward Chandler for being bad.

    (I know this is a simplification. Chandler creates “assist opportunities” for himself, setting good picks, etc. … but you can only get so many of this. And i believe we have witnessed how the Knicks kept Chandler involved in a lot of sets, but only a few of them produced a Tyson smash)

  125. I think i got Chandler’s unassisted percentage wrong. Do you know somewhere to look it up?

  126. So, somehow, WP has decided that it has to penalize players that are better than Chandler, and reward Chandler for being bad.

    (I know this is a simplification. Chandler creates “assist opportunities” for himself, setting good picks, etc. … but you can only get so many of this. And i believe we have witnessed how the Knicks kept Chandler involved in a lot of sets, but only a few of them produced a Tyson smash)

    But Chandler isn’t bad. He’s good at playing within himself, and very good at sucking defenses far from other much better outside shooters. Win-win!

    In related news: Kyle Korver (who has the same effect of Tyson, only outside-in, without rebounding, stealing, assisting or whatever much) has posted a .231 WP48 since Budenholzer took the reins. So sometimes is the system that makes the player (never cracked .200 before). Tyson has very few elite skills, and if he was made to be a post presence his WP48 would surely plummet. But if you use him in the right ways, he is a force of nature because he distorts the center of gravity on the field both on offense and on defense.

  127. I understand there could be a case for high usage players to see their efforts rewarded more by advanced stats, at least on offense (shouldering the load for LeBron is a much, much harder task than for, say, Hassan Whiteside).

    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.):

    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

    Hey, it seems it’s doing good work!

  128. Let’s try FWP48 for our beloved Knicks. Here are the guys who posted positive scores (Knicks stint only).

    Shved: .135
    Stat: .077
    Prigs: .069
    Cole: .061
    Melo: .031
    Shane Larkin: .016
    Jose: .001

    All the other guys post negative values. The most notable ones:

    Bargs: -.202 (worse than Kobe!)
    Ricky Ledo: -.197 (worse than Kobe, better than Bargs!)
    Jah: -.115
    Timmy: -.138

    Welp.

  129. I like the FWP48. We have above average point guard play and below average scoring guard play.

  130. “But Chandler isn’t bad. He’s good at playing within himself, and very good at sucking defenses far from other much better outside shooters.”
    “he is a force of nature because he distorts the center of gravity on the field both on offense”
    This is utter nonsense. He is no better at sucking defenders anywhere than a whole host of big men in the league. There is myth out there that he has some amazing finishing skill that no other player in the league has. He shoots at about 70% near the rim. So do a bunch of other big men including, but not in any way limited to, Whiteside, DeAndre Jordan, Amare, David Lee, Blake Griffin, the brothers Gasol and Marcin Gortat. His shooting numbers look gaudy because he absolutely refuses to take any shot that doesn’t involve dropping the ball downward. Here’s another thing — people on this site call DeMarcus Cousins a turnover machine. You know who has a higher turnover percentage than Cousins? Tyson Chandler. He is a true turnover machines since he turns the ball over one out of every six possessions, which is outrageous since he never dribbles, passes or tries to create his own shot. (By the way, his turnover rate is 50% higher than Melo’s.) He is asked to do one thing — catch the ball up high near the basket and drop it into the basket, yet he flubs that role 16% of the time. He is a good, but extremely limited, player. He can’t dribble, pass or shoot and is capable of doing one thing, and one thing only, on offense.

  131. Johnno, Tyson has a high TOV% because he shoots so infrequently. TOV% is 100 * TOV / (FGA + 0.44 * FTA + TOV). Say you have two guys who touch the ball 10 times in a game. Both turn the ball over twice. Player A shoots one time. Player B shoots two times.

    Player A TOV is 100 * 2/(1+2)=66

    Player B TOV% is 100* 2/(2+2)=50

    Which guy turned the ball over more frequently?

    Cousins and Chandler both had a TOV% of about 16 this season, but Cousins averaged about 3 more turnovers per game.

    Per 36 minutes, Cousins averaged about 13 more PPG than Tyson. That’s a lot. However, he took about 12 more shots a game than Tyson, while giving the ball back to the other team 3 more times and getting 1 fewer offensive rebound. A roughly average team is going to turn those 16 possessions into 16 points, which means Tyson’s team would score 3 more points than Boogies, which is why Tyson is better for your offense than Demarcus Cousins, despite Boogie’s edge in both points scored and skill possessed.

  132. The (rather massive) assumption in those numbers being that everything else is equal between having Boogie on the floor vs. having Tyson.

  133. “Tyson has a high TOV% because he shoots so infrequently. ”

    I mean thats not really a defense is it? That just means that he has no reason to turn the ball over at all if its so infrequent. You cant take all the credit for not shooting, then use it as an excuse.

  134. I mean thats not really a defense is it? That just means that he has no reason to turn the ball over at all if its so infrequent. You cant take all the credit for not shooting, then use it as an excuse.

    Tyson doesn’t get credit for not shooting. He gets credit for making shots at a very high efficiency. He doesn’t get credit for not turning the ball over-he loses credit for his turnovers. . .he just doesn’t turn it over that often.

  135. Chandler hardly ever touches the damn ball except on lobs and rebounds. The vast majority of his “passes” consist of handing the ball to a guard after getting a defensive rebound or tapping the ball to a teammate on an offensive rebound. Ballhog Carmelo Anthony has an assist rate that is almost 4 times Chandler’s. How many turnovers do you think he would have if he actually tried to dribble or throw a pass that travelled more than 2 feet? It’s interesting that the Chandler fans are willing to cite all of his good advanced stats (as if there is no context ) yet, when the advanced stats stink, context is the excuse. He is “efficient” on offense because he can’t do — and to his everlasting credit doesn’t even attempt to do — what most pro basketball players can do and actually do on the court — dribble, pass and shoot. As I said in my prior post, the one elite offensive skill that he has (dropping the ball in the basket 3 times a game) is not the slightest bit unique — loads of tall athletes can do it and actually do do it.

  136. It reads to me like TOV % is the ratio of TO to FGA + % of FTA. Same TO% does not = same TOs/g or /36.

  137. Whether or not a skill is unique doesn’t fucking matter. What Tyson does on the court on offense generally helps his team score more points than the other team. Even though he can’t pass like Carmelo, even though he would turn the ball over a lot if he had to dribble and create shots off the bounce, even though he does nothing but drop the ball in the basket which loads of other tall athletes do-who gives a shit? Everything you do on a basketball court only matters in how much it helps your team score more points than the other team. Tyson scores really efficiently and generates additional possessions for his team. Those things help his team outscore the other team more often than not. That’s what matters.

  138. As I said in my prior post, the one elite offensive skill that he has (dropping the ball in the basket 3 times a game) is not the slightest bit unique — loads of tall athletes can do it and actually do do it.

    And yet only one person has the all time seasonal TS% record.

  139. Whether or not a skill is unique doesn’t fucking matter. What Tyson does on the court on offense generally helps his team score more points than the other team. Even though he can’t pass like Carmelo, even though he would turn the ball over a lot if he had to dribble and create shots off the bounce, even though he does nothing but drop the ball in the basket which loads of other tall athletes do-who gives a shit? Everything you do on a basketball court only matters in how much it helps your team score more points than the other team. Tyson scores really efficiently and generates additional possessions for his team. Those things help his team outscore the other team more often than not. That’s what matters.

    (i can’t believe we’re having this discussion again)

    If you arbitrarily think of basketball as some summation of dribbling, passing, defending, and scoring, then all 4 of these things matter. Tyson can do 2 out of the 4 very well, but that means that someone else has to do the other two. It’s nothing against Tyson, it’s just that he abdicates responsibility for 50% of what basketball is about. And again, it’s good that he doesn’t try to do things he’s bad at, but that doesn’t change the fact that those things need to be done by someone. Dribbling and passing are also the things that tend to lead to most turnovers, so Tyson’s TOV% or TO/g are not really comparable to someone who has a completely different set of responsibilities.

    There’s definitely value in not doing what you suck at – we saw what happens when players try to do things outside their limits – ie. Dalembert in particular.

  140. Tyson can do 2 out of the 4 very well, but that means that someone else has to do the other two. It’s nothing against Tyson, it’s just that he abdicates responsibility for 50% of what basketball is about.

    And that would matter if it hurt his teams overall performance. But it seems that it does not. So it does not matter.

  141. There is no correlation between a team (1) having a player who takes very few shots relative to his teammates and (2) an overall drop in efficiency among his teammates. If there were, the metric would account for that.

    This is a Wins Produced fallacy.

    Even Berri comes close to admitting that if you built a team with Chandler, Fields, Calderon, etc.. the model would break down at that extreme because they wouldn’t be able to score efficiently. You need skilled scorers. The fact that it breaks down at real extremes tells you usage is an issue.

    The reason why the usage/efficiency issue doesn’t show up in Berri’s studies is that he could not control for player development (players increase their usage as they improve) and because coaches aren’t foolish enough to combine too many non scorers together often so you could actually test the impact.

  142. @175 – I don’t hate Chandler at all. To the contrary, I like him a lot as a player (or at least I did until he quit on the team last year). I think he is a valuable player on a good team and I am being serious when I say I appreciate that he doesn’t try to do what he is incapable of doing, because there is real value in not doing dumb things on the court. I just think that it’s silly to claim that he is a top 10 in the league type of player.

  143. And that would matter if it hurt his teams overall performance. But it seems that it does not. So it does not matter.

    If you played Tyson Chandler, Deandre Jordan, Andre Drummond, Hassan Whiteside, and Rudy Gobert as a lineup, would that be a good team or a bad team?

  144. Frank – why would anyone start 5 centers? It would be a 7 second violation or TO every time. What about Nate, Jamal, Carmelo, Nick Young and Kobe how would they do? I can’t believe this place has devolved into a version of automated phone heel “if you think Melo is a volume scorer press 1” “if you want to talk about the Melo trade press 2” “if you think Tyson is an overrated cherry picking efficiency stat padder press 3” “if you want to rail against WOW, PER, WP/48, WS/48, TS%, TO% … press 4” “if you want to rehash Jeremy Lin Press 4” and so on and so on.

  145. The only reason to bring up extremes (ie. starting 5 centers) is that DRed is saying that Chandler’s inability to dribble or pass doesn’t hurt the team. If that’s the case, why bother with dribblers or passers at all?

    The fact that it’s absurd to even consider that lineup tells you that yes, dribbling and passing are important skills that someone has to have in order to mount a reasonable offense. And the fact that Chandler can’t do it means someone else has to. And so he is abdicating responsibility for some large percentage of what the game is about.

    it’s not so different from where Rondo finds himself now (and even at the height of his abilities)- he dribbles well, passes well, and can defend when he wants to – but the fact that he can’t shoot at all makes life harder on those around him. You could theoretically construct a team to hide that deficiency (ie. give him 3 hall of famers to play with), but without the right mix it falls apart.

    Tyson is a far superior player to Rondo, especially now, but the situations are not so different.

  146. @180

    I agree.

    Decades of observing on court results lead coaches to understand that the best results are produced by creating a team with a balanced attack of various skills (very few players have all of them). That is not to say that we can’t capture some of the values better with advanced stats, but when the implication of a model is that player “X” is a top player because he rebounds well for the position, doesn’t turn the ball over much, and is smart enough to not shoot shots he can’t make too often, there’s often a problem. Lack of offensive skill is a negative, not a positive. It doesn’t become a positive just because the player made 2 of 3 dunks for an above average efficiency. Wins Produced is probably correct to assume that there is some ability for players to adjust their usage with minimum impact to team or personal efficiency, but it is incorrect to think that usage doesn’t matter at all. To think otherwise is to think everyone that has been playing, observing, coaching, and studying the game for decades can’t tell that Landry Fields is actually a great player. No, it’s that the model cannot tell he’s not.

  147. I think people are significantly overestimating the impact that a single Chandler’s lack of “skills” has on a teams efficiency. Increasing the amount of Chandlers on the floor means the “negative” impact doesn’t increase linearly–five Chandlers is so much different from a one chandler lineup that it doesn’t make sense to compare the two unless you’re into rhetorical sleight of hand. Of course a team of five Chandlers would be bad because of poor ball handling and shooting, but one Chandler doesn’t negatively affect a teams play options in a commensurate way (it’s not the negative value of 5 chandlers divided by 5–its far less. 5 Chandlers is an edge case that should have little to no bearing on a team with one Chandler.

  148. Generally speaking using a strictly theoretical counterfactual to disprove an empirical assertion does pretty much zero argumentative work because of course you can arbitrarily construct an edge case scenario in which the assertion is unsound.

    To me it seems like Chandler doesn’t seem to impact his teams negatively on offense–hes been on plenty of teams with solid offensive efficiency numbers and scores his team points in an efficient manner–im on my phone but im sure his offensive on/off numbers are favorable to him. All this adverting to interaction effects just seems a bit silly to me.

  149. You should use stats to describe, not prescribe. That’s my attitude. I don’t care if Tyson Chandler were a bad basketball player, if the game were to revolve around taking left-handed, blind-folded hook shots. He is a tremendously productive basketball player, and it’s up to coaches to replicate whatever he’s been doing for the last five years and stick fuckin’ to it.

    Hmm. What I’m getting from this — as well as the earlier “eye test plus advanced stats” answer — is that these stats aren’t necessarily ways to evaluate players in a vacuum. That is, we can’t say if Tyson is a better player than Melo is — just that he’s more productive given his role in the offense. Therefore, it’s the coach’s job to determine where his players are most productive — in what position, what role, etc? I’m thinking this line of thinking can apply to arguing for more #MeloAtThe4.

    The impression I’m getting, therefore — and correct me if I’m wrong — is that the advanced metrics we’re using are inherently flawed in terms of banging out a conclusive “Top 15 players in the NBA” list. What we see based on these stats is “Who contributes the most to team success” — and this list is intrinsically affected by the team’s situation.

    Thanks everyone! :)

  150. Tyson being unable to shoot or generate his own offense shifts responsibility for roughly 5 extra FG attempts to his teammates (compared to an average center). That is less than one shot per game per man (assuming teams play around 8-9 guys a night). If you think that creates an increase in usage such that the other players efficiency drops enough to overcome the significant increase in possessions Tyson’s efficiency and offensive rebounding create than I would love to see how you came to that conclusion, because it seems nonsensical. For a player like Tyson to have a negative impact on the team’s offense relative to an average center, the team would have to have the worst offense in basketball by a huge margin. Lack of offensive skill is a negative that is massively outweighed by the positive effects of having a ridiculously good TS% and generating lots of offensive rebounds while not turning the ball over. Plus he sets good picks.

  151. “Lack of offensive skill is a negative that is massively outweighed by the positive effects of having a ridiculously good TS%”

    Not sure that this goes together. His TS% is caused by his lack of skill isnt it? I dont think its something that balances his lack of attempts.

  152. I’m thinking this line of thinking can apply to arguing for more #MeloAtThe4.

    I’m for anything that argues for Melo at the 4. ;)

  153. There’s something major getting lost in the “skills” argument against Tyson Chandler. It’s true that someone like Demarcus Cousins has shown a greater ability to do things like hit mid range jump shots and score in isolation, so if you want to say he’s more skilled in those respects fine…but why are we ignoring the purely physical skills in which Tyson is superior? Cousins ain’t exactly catching lobs or getting nearly as many offensive rebounds. These are both things that have to do with the “skills” Tyson excels at (namely being really athletic), the only difference is they have a much greater correlation with winning.

    So DeMarcus Cousins is more “skilled” in some respects, and Chandler is more “skilled” in others that actually lead to wins. I use parenthesis because I find the term rather devoid of meaning (if your skills are detrimental to basketball games being won, like Cousins’ propensity for jump shots likely is, should we still categorize them that way?).

  154. Roughly a decade ago, when he was just writing for Basketball Prospectus, John Hollinger wrote about how players like STAT were being underrated when it came to their skills in the pick and roll, because STAT made it look so easy. I think that’s being done to Chandler (and players of his ilk) as well. It is a unique skill to be able to be into position and throw down dunks. If it were easy, more tall players would do it. They don’t. Because it isn’t easy. Chris Paul, Jason Kidd, Raymond Felton and Jameer Nelson – the point guard side of the pick and roll has changed dramatically in skill level and yet Chandler’s side remains highly effective (his only not quite as amazingly efficient season over the past nine seasons was his final year in New Orleans, where he was hurting so badly that he was actually traded and the trade was rescinded due to how injured he was – and even then he was quite efficient, just not as efficient as the other seasons).

Comments are closed.