Statistical Analysis. Humor. Knicks.

Wednesday, September 17, 2014

Knicks Morning News (Wednesday, Dec 19 2012)

  • [New York Times] Advertising: Chris Paul to Star in State Farm Insurance Ads (Wed, 19 Dec 2012 06:48:03 GMT)
    Chris Paul of the Los Angeles Clippers will appear in a campaign for State Farm Insurance designed to lure younger customers.

  • [New York Times] Jazz 92, Nets 90: Nets’ Deron Williams Loses Again to Former Team (Wed, 19 Dec 2012 06:37:21 GMT)
    Deron Williams had 14 points and 5 assists in what qualified as another substandard performance, while the Nets lost for the seventh time in nine games.

  • [New York Times] Knicks’ Stoudemire and Anthony Move Closer to Playing Together (Wed, 19 Dec 2012 06:37:21 GMT)
    Amar’e Stoudemire said he would not object to coming off the bench rather than reassuming his role as the starting power forward, a position currently filled by Carmelo Anthony.

  • [New York Times] N.B.A. Roundup: Nuggets Hold Off Spurs’ Rally (Wed, 19 Dec 2012 06:30:07 GMT)
    Danilo Gallinari scored a season-high 28 points and outplayed Tim Duncan down the stretch as the host Denver Nuggets beat the San Antonio Spurs, 112-106, on Tuesday night.

  • [New York Times] Lin Remains a New York Story (Wed, 19 Dec 2012 06:30:07 GMT)
    Jeremy Lin, now with the Houston Rockets, was, in a sense, perfect for a city that has long relished the ethnic fan experience. But the Knicks turned their backs on that.

  • [New York Times] New York High School Basketball Players Migrate to New Jersey (Wed, 19 Dec 2012 06:20:07 GMT)
    New Jersey has become more of a talent hotbed than New York City, coaches say, with some top high school basketball players departing the boroughs.

  • [New York Times] Lakers Hand Bobcats 12th Straight Loss, 101-100 (Wed, 19 Dec 2012 06:17:50 GMT)
    Kobe Bryant scored four of his 30 points in the final 1:26 and the Los Angeles Lakers overcame an 18-point third-quarter deficit to beat Charlotte 101-100 on Tuesday night, handing the Bobcats their 12th straight loss.

  • [New York Times] Warriors Hold Off Hornets 103-96 in Return Home (Wed, 19 Dec 2012 06:17:45 GMT)
    David Lee had 26 points and nine rebounds, Klay Thompson finished with 19 points and the Golden State Warriors brought their surprising road run home with a 103-96 victory over the struggling New Orleans Hornets on Tuesday night.

  • [New York Times] Heat Tame Timberwolves Despite Huge Rebounding Gulf (Wed, 19 Dec 2012 05:29:35 GMT)
    The Miami Heat overcame a 28-point rebounding differential to beat the Minnesota Timberwolves 103-92 on Tuesday with Dwyane Wade top scoring with 24 points.

  • [New York Times] Gallinari’s 28 Points Lifts Nuggets Past Spurs (Wed, 19 Dec 2012 05:06:24 GMT)
    Danilo Gallinari scored a season-high 28 points, outlasting Tim Duncan down the stretch and leading the Denver Nuggets over the San Antonio Spurs 112-106 on Tuesday night.

  • [New York Times] Off the Dribble: Cavs’ Varejao Is Invaluable and Tradeable (Wed, 19 Dec 2012 04:55:40 GMT)
    Anderson Varejao has never been better. He could help the Cavs make the playoffs — or rebuild for the future.

  • [New York Times] Off the Dribble: Raymond Felton and Jeremy Lin Adjust to New Teams (Wed, 19 Dec 2012 04:39:27 GMT)
    While Raymond Felton has made himself at home in his second stint with the Knicks, Jeremy Lin and the Rockets are building as they go.

  • [New York Times] Mayo, Kaman Help Mavericks End 3-Game Skid (Wed, 19 Dec 2012 04:35:42 GMT)
    O.J. Mayo scored 26 points, Chris Kaman added 20 and the Dallas Mavericks snapped a three-game losing streak by beating the Philadelphia 76ers 107-100 Tuesday night.

  • [New York Times] Wade, James Carry Heat Past Timberwolves, 103-92 (Wed, 19 Dec 2012 04:14:44 GMT)
    The Miami Heat were outrebounded by 28, matching the second-largest margin in team history. They finished with only 24 boards, matching the second-lowest total in any game over the franchise’s quarter-century of existence.

  • [New York Times] Jennings Helps Bucks Beat Pacers (Wed, 19 Dec 2012 03:57:04 GMT)
    Brandon Jennings scored 13 of his 34 points in the fourth quarter and Monta Ellis had 19 points, leading the Milwaukee Bucks to a 98-93 win over the Indiana Pacers Tuesday.

  • [New York Times] Deng, Boozer Score 21 as Bulls Beat Celtics 100-89 (Wed, 19 Dec 2012 03:51:05 GMT)
    Luol Deng and Carlos Boozer scored 21 points apiece, Joakim Noah had a triple-double and the Chicago Bulls beat the Boston Celtics 100-89 on Tuesday night.

  • [New York Times] Williams Scores 24, Hawks Beat Wizards in OT (Wed, 19 Dec 2012 03:47:51 GMT)
    Lou Williams scored a season-high 24 points and Josh Smith had 17 points and 13 rebounds to lead the Atlanta Hawks to a 100-95 overtime win over the Washington Wizards on Tuesday night.

  • [New York Times] Jazz Rally, Then Hold On to Beat Nets 92-90 (Wed, 19 Dec 2012 03:08:59 GMT)
    Mo Williams scored 19 points, Al Jefferson had 16 points and 11 rebounds, and the Utah Jazz beat the Brooklyn Nets 92-90 Tuesday night despite nearly throwing away the victory.

  • [New York Times] Calderon Leads Raptors Past Cavaliers 113-99 (Wed, 19 Dec 2012 02:54:04 GMT)
    Jose Calderon scored a season-high 23 points, and the Toronto Raptors won on the road for the second time this season by beating the Cleveland Cavaliers 113-99 on Tuesday night.

  • [New York Post] Woodson focused on defense, not minutes (Wed, 19 Dec 2012 02:54:24 -0500)
    Mike Woodson allowed himself to daydream for a moment yesterday about a day when all of his key players are healthy and the Knicks have a full roster for him to tinker with.
    “It’s going to be interesting to see where we go once everybody is in place,â? he…

  • [New York Post] Healing Amar’e â??totally open’ to coming off bench for Knicks (Wed, 19 Dec 2012 02:56:54 -0500)
    The $100 million man is willing to become a sixth man if necessary.
    Amar’e Stoudemire, with $65 million and three years left on his $100 million contract, said yesterday he will do “whatever it takes,â? including accepting a role off the Knicks’ bench, to win and avoid disrupting team…

  • [New York Post] Carmelo not caught up in rush hour (Wed, 19 Dec 2012 03:13:35 -0500)
    Carmelo Anthony wants to take the smart route. And the smart route is not rushing back.
    So it’s possible Anthony will miss a third straight game because of a sprained left ankle tonight when the Knicks entertain the Nets â?? for the first time as the Brooklyn Nets â?? at Madison…

  • [New York Post] Melo looking for an edge in rivalry (Wed, 19 Dec 2012 01:44:39 -0500)
    Round 3 is at the Garden this time. The Knicks-Nets Battle of the Boroughs has gone through two legs at Brooklyn’s Barclays Center with the teams managing a split in two wildly exciting, close battles. The Nets survived 96-89 in overtime on Nov. 26. Then, on Dec. 11, the…

  • [New York Daily News] Amar’e cautious about return date with Knicks (Wed, 19 Dec 2012 06:54:12 GMT)
    There is still a chance Amar’e Stoudemire might be home for Christmas â?” and not with the Knicks in Los Angeles to face the Lakers. Stoudemire cautioned that he’s still “not totally there yetâ? while downplaying any imminent return to the lineup.

  • 60 comments on “Knicks Morning News (Wednesday, Dec 19 2012)

    1. d-mar

      Weird game in Miami last night, the T-Wolves outrebounded the Heat 53-24 and lost by 11. In the process, however, they committed 19 turnovers, which is a killer vs. a team like the Heat.

      I think that’s one of the reasons we’ve had success vs. Miami so far, our ability to take care of the ball. Take away their transition baskets, and they’re a more ordinary team.

    2. iserp

      The Honorable Cock Jowles: Who gives a shit about PER? And who cares that Martin is +.040 or so in WP48? It’s been twenty games and Harden is shooting 59% TS on 22 FGA/48. Gloat at the end of the year, big guy.

      Wow, wow, wow. James Harden is playing like most people expected he would be playing, Kevin Martin is playing like most people expected he would be playing. PER agrees with that those players have met expectations.

      20 games into the season (not a bad sample after all), Kevin Martin is .120 WP48 higher (as per thenbageek.com) than last year and James Harden is 0.090 lower. That’s the difference between an average player and a star. I think it is meaningful. I’ve always criticized the lack of predictive power of WP48, and that the consistency of numbers across seasons are an artifact of the NBA career of players (players usually stay in their teams, and when they change teams, coaches tend to use them in the same role they used to have). There are few opportunities to analyze this, let’s talk it.

      Both players have logged heavy minutes, haven’t suffered injuries, and their teams/coaches seem happy with their acquisition. This deserves at least a “JH and KM are an outlier because blah blah blah…”

    3. DS

      d-mar: I think that’s one of the reasons we’ve had success vs. Miami so far, our ability to take care of the ball. Take away their transition baskets, and they’re a more ordinary team.

      Second. That will be an important consideration when we can’t count on 18 3-pointers per game going in during a playoff series.

    4. DS

      I enjoyed Mr. Cavan’s piece on Sideshow Bob, since I’m one of those people whose favorite waste of time is the ESPN Trade Machine.

      The idea of him going to OKC seems to make a lot of sense, since rebounding is the closest thing the Thunder have to a weakness and they could send to Cleveland some of their haul from the Harden deal. But you have to figure if there was a deal to be made, Presti would have made it. They’re prob. insistent on sending Perkins back which is prob. a deal breaker.

    5. DS

      Also, I’m not afraid to point out the obvious – or to triple post for that matter; the Knicks are not rebounding well at all so if the Cavs are interested in a package around Shump, JR?, and picks for Varejao, it might be something to consider; crowded front court be damned.

    6. iserp

      DS: Also, I’m not afraid to point out the obvious – or to triple post for that matter; the Knicks are not rebounding well at all so if the Cavs are interested in a package around Shump, JR?, and picks for Varejao, it might be something to consider; crowded front court be damned.

      I wouldn’t consider it. I think it is more important for us to have ballhandlers in the 2nd unit than it is to play an offensive challenged C next to chandler.

      Once Amare returns, we will be able to play bigger lineups and the rebounds will tick a little bit upwards.

    7. thenamestsam

      d-mar:
      Weird game in Miami last night, the T-Wolves outrebounded the Heat 53-24 and lost by 11. In the process, however, they committed 19 turnovers, which is a killer vs. a team like the Heat.

      I think that’s one of the reasons we’ve had success vs. Miami so far, our ability to take care of the ball. Take away their transition baskets, and they’re a more ordinary team.

      Yeah, that was one of the most entertaining non-Knicks games I’ve seen all year. Fascinating contrast of styles. Interestingly, the Heat seem to have really worked out a lot of their defensive problems by resorting to going big again. They couldn’t get a rebound to save their life last night, but just in terms of how they defended initial possessions they were really great. Even though the offensive rebounds generated a ridiculous 58 points in the paint Miami still held them to 43% shooting and generated those 19 turnovers. But to fix the defensive problems they’ve mostly given up the small-ball identity that they found in the playoffs last year. It’ll be interesting to see if they switch back to small-ball come playoff time.

    8. Frank

      iserp: 20 games into the season (not a bad sample after all), Kevin Martin is .120 WP48 higher (as per thenbageek.com) than last year and James Harden is 0.090 lower. That’s the difference between an average player and a star….I’ve always criticized the lack of predictive power of WP48, and that the consistency of numbers across seasons are an artifact of the NBA career of players (players usually stay in their teams, and when they change teams, coaches tend to use them in the same role they used to have). There are few opportunities to analyze this, let’s talk it.

      A few threads back, I literally took every single significant player move this offseason and compared their WP48 in their new surroundings with their 2-year averaged WP48 for their old teams — and found that only 6 of the 20 players analyzed had a WP48 even in the same zip code this year as their 2 year average for their old team. Obviously it was only a 1/4 season sample in the new place, and maybe all those differences would even out over a bigger sample. But all you have to do is think about all the big player moves – Deron Williams has sucked since coming to the Nets – was a top 3 WoW PG in Utah, and now is fighting with Jamaal Tinsley to be in the top 30 PGs in WP48. Amare’s efficiency fell off a cliff once he left Nash. David Lee turned from greatest WoW player in the world in 2 steps – when his usage increased in NY, and then again when he became a WP48 0.05 guy in GSW (to be fair, he is back up to 0.177 this year). Zach Randolph went from a guy WoW hated to a top 3 PF once he went to Memphis, in his 9th season no less.

      I appreciate WoW because it greatly values ball possession (and not turning the ball over) and also shooting efficiency, and it clearly and correctly values guys like Kidd, Ronnie Brewer, etc. more than “traditional” “advanced” stats do. continued….

    9. JC Knickfan

      yellowboy90: What sucks is that the long two’s are his best shot % wise.
      Also copeland plays no defense and doesn’t look quick enough to stay with 3s or big and strong enough to hold up against 4s.

      The team is 18-6 so I think hard argue to the changing the status quo.

      My thinking is Copeland could provide better offensive output by taking away Felton shoots.

      On defense the Copeland/Melo frontcourt, I think you ask Melo to take better player. You have look match before saying Copeland will be destroyed 3 or 4’s. Next game is the Nets. Gerald Wallace is better offensive and I would have Melo cover him. Copeland would cover Humphries whose average 7.4 ppg. I wouldn’t be too scared Nets trying post Copeland up, but more concern if Copeland could keep Humphries off the boards.

      But base his last start, I would like to see Copeland get another chance.

    10. Frank

      My issue with WoW is, for the most part, the same as Juany8′s.

      Somehow defense is a team game where all 5 players get the same credit. But on offense, players are 100% responsible for their own stats?

      I went through this reasoning before, and I’d love THCJ or Owen to comment on it.

      - PF/C generate more wins than guards. If so, why don’t teams field a whole team of PF/Cs? Why not line up Chandler, Varejao, Andre Drummond, Reggie Evans, and Tiago Splitter?
      - if your answer is that, duh, you NEED guards because centers can’t do the stuff that guards do, you’ve basically agreed that there are serious and inseparable interactions between the various positions on the floor.
      - once you’ve agreed that there ARE interactions, then you can’t say with any certainty that Kenneth Faried is the greatest player in the world, because he only succeeds at his job because other players on the team allow him to do only what he does best.
      - some players are more dependent than others on their teammates for their own production. Some, like Varejao, basically had indistinguishable stats pre- and post-LBJ’s departure from Cleveland. Good for him. Others, like Mo Williams, basically imploded once he actually had players defending him rather than triple-teaming Lebron.
      - on our own team, how do we explain Landry Fields’s collapse after the Melo trade if Landry and only Landry was responsible for his own stats?

    11. Juany8

      iserp: Wow, wow, wow. James Harden is playing like most people expected he would be playing, Kevin Martin is playing like most people expected he would be playing. PER agrees with that those players have met expectations.

      20 games into the season (not a bad sample after all), Kevin Martin is .120 WP48 higher (as per thenbageek.com) than last year and James Harden is 0.090 lower. That’s the difference between an average player and a star. I think it is meaningful. I’ve always criticized the lack of predictive power of WP48, and that the consistency of numbers across seasons are an artifact of the NBA career of players (players usually stay in their teams, and when they change teams, coaches tend to use them in the same role they used to have). There are few opportunities to analyze this, let’s talk it.

      Both players have logged heavy minutes, haven’t suffered injuries, and their teams/coaches seem happy with their acquisition. This deserves at least a “JH and KM are an outlier because blah blah blah…”

      What’s even funnier is that last year had a higher WP48 than Kevin Martin, OJ Mayo, and Kevin Martin combined. Now that the circumstances for those 3 players noticeably changed, all 3 have close to the same WP as Harden. Even more amusing? After crapping the bed all of last year, Landry Fields STILL was close to a star in terms of WP, with a .170 WP48 for last year. In fact going by his first 2 years, Toronto got a steal with the contract they gave him. I wonder how many players were superstars as rookies and out of the rotation of a shitty team in 2 years despite pristine health.

    12. iserp

      Frank, I agree with most your points. I want to add something, WoW is fundamentally rooted in the importance of the possesion. However, which players nets you possesions to your team is not only a matter of ability, but of role. If we could parametrize a basketball game perfectly, we would have to parametrize separatedly the abilities of the players (which don’t change from team to team, coach to coach, game to game, …) from the role you play in the team. If you use a single stat to measure a player, it is probably going to mix the information that comes from the abilities of the player and the role he takes. If you compare players with similar roles… then WP48 (and most ‘advanced stats’) will be useful. But WP48 is specially sensitive to changes in roles, and makes it useless in lots of common situations. PER at least takes usage into account (which would be one of the parameters of your ‘role’), so it will ‘work’ in more situations than WP48.

      Frank: – on our own team, how do we explain Landry Fields’s collapse after the Melo trade if Landry and only Landry was responsible for his own stats?

      I have to say that something must have happened to Landry Fields. His FT% plummeted. That’s not normal. Free throws are one of the “purest” abilities you have. It doesn’t matter the coach, team, role, you will shoot more or less the same percentage (with some deviance based on the number of FT you shoot, specific training, etc…). Let’s say FT% is directly related to your ability to hit open jumpers (with some coefficient to account for distance). Landry Fields was valuable to the offense when he could hit open 3s, he can’t do much else … so he is worthless now. I hope he returns to his older form (No way he is worth his contract either way, but he is an OK role player and a nice guy)

    13. Frank

      iserp: I have to say that something must have happened to Landry Fields.

      Yes, maybe that was a bad example. Clearly there was some sort of a Landry-specific mental block. Seriously, some of his shots never even got above the level of the rim. I’ve literally never seen an NBA shooting guard miss free throws by as much as he missed them by (feet!).

    14. Nick C.

      Technically Fields is injured. I thought I read he was having wrist or some such surgery. Be that as it may he was something of a nominal starter so I am splitting hairs.

      As for many of the WP/48 etc. don’t they come up with values based on how they correlate on a team level. That was my understanding. If so then even if the data is solid on teams, how do you fairly and accurately parcel it out to individuals. If one is just making guesstimates then is it really that reliable. (an example A plays great defense contests a shot that is missed and B grabs it uncontested-if I understand correctly B gets +1 rebound + 1/5 the decrease on efg% or whatever measure they use for defense, A gets the same 1/5 defense credit)

    15. Juany8

      Nick C.:
      Technically Fields is injured. I thought I read he was having wrist or some such surgery. Be that as it may he was something of a nominal starter so I am splitting hairs.

      As for many of the WP/48 etc. don’t they come up with values based on how they correlate on a team level. That was my understanding. If so then even if the data is solid on teams, how do you fairly and accurately parcel it out to individuals. If one is just making guesstimates then is it really that reliable. (an example A plays great defense contests a shot that is missed and B grabs it uncontested-if I understand correctly B gets +1 rebound + 1/5 the decrease on efg% or whatever measure they use for defense, A gets the same 1/5 defense credit)

      That’s exactly the problem Nick, things like offensive efficiency and rebounds correlate really well with wins at the team level, but WP just assumes that you can directly translate those stats to the team level. Since they also assume there is no correlation between efficiency and usage, they basically assume that players take the number of shots they take based purely on choice, either by the coach or player. As we know that couldn’t be further from the truth.

      Berri assumes Chandler only takes shots at the paint because he’s a brilliant player, common basketball sense tells us he only takes shots at the rim because that’s the only kind of shot he can take at anywhere near a respectable level. Tim Duncan shoots a higher percentage at the rim than Chandler, but because he’s “dumb” enough to take shots more than 2 feet from the rim, he’s somehow a worse offensive player than Chandler.

      WP assumes that each player is entirely responsible for his own box score stats, but that the team collectively is responsible for the defense, which isn’t in a box score. That’s not logic that’s convenience

    16. DS

      iserp: I wouldn’t consider it. I think it is more important for us to have ballhandlers in the 2nd unit than it is to play an offensive challenged C next to chandler.

      Once Amare returns, we will be able to play bigger lineups and the rebounds will tick a little bit upwards.

      I agree, though to nitpick, I prob. wouldn’t call Varejao offensively challenged for the reasons Jim Cavan gives… a smarter move might be to see if they have redundancies when STAT and Shump are back and deal somebody for a rebounding “specialist.”

    17. Frank

      ruruland:
      Gosh, it’s almost like Owen and Jowles disappeared again. Wonder what that’s about.

      That’s the thing that disappoints me the most about all the fights or discussions we have here. There are lots of categorical statements of Berri-oid philosophy, and certainly lots of insults thrown around (not by Owen for the most part), but very little constructive back-and-forth when posters have actual substantive questions/issues about WoW.

    18. Nick C.

      I think basketball just doesn’t lend itself to individualized ratings as baseball. Baseball what you do at the plate is essentially your own doing (other than maybe altering your approach based on baserunners). In the field more or less the same thing. Basketball not so much.
      An off-target pass from a spazz who passes up the shot, less than accurate PG etc. may result in a heavily guarded force while a well thrown pass that can be caught and shot or dribbled in one smooth motion might result in a clean look. That being said advanced numbers like EFG%, TSP and others have helped matters so that never again will we be stuck valuing people based on PPG.

    19. ruruland

      Frank: That’s the thing that disappoints me the most about all the fights or discussions we have here.There are lots of categorical statements of Berri-oid philosophy, and certainly lots of insults thrown around (not by Owen for the most part), but very little constructive back-and-forth when posters have actual substantive questions/issues about WoW.

      This

    20. DRed

      I’ll get to this when I’m done covering the game I’m at.

      You think I’m going to fall for this? We both know how it’ll end.

      I’ll get to this, too. You’re off. See search button if you want to know where I’ll go.

      Gosh, it’s almost like Owen and Jowles disappeared again. Wonder what that’s about.

      Those are the last 4 posts from ‘why can’t we have a constructive back-an-forth’ ruru.

    21. The Honorable Cock Jowles

      Frank: That’s the thing that disappoints me the most about all the fights or discussions we have here.There are lots of categorical statements of Berri-oid philosophy, and certainly lots of insults thrown around (not by Owen for the most part), but very little constructive back-and-forth when posters have actual substantive questions/issues about WoW.

      EXPLAIN HOW THERE’S AN OUTLIER!

      EXPLAIN HOW DEFENSE CAN BE DISTRIBUTED AMONG TEAMMATES! (Then use Synergy stats, even though they attribute 30% of plays to “team defense” themselves.)

      EXPLAIN HOW [PLAYER X] COULD BE RANKED THAT HIGH!

    22. ruruland

      DRed:
      I’ll get to this when I’m done covering the game I’m at.

      You think I’m going to fall for this? We both know how it’ll end.

      I’ll get to this, too. You’re off. See search button if you want to know where I’ll go.

      Gosh, it’s almost like Owen and Jowles disappeared again. Wonder what that’s about.

      Those are the last 4 posts from ‘why can’t we have a constructive back-an-forth’ ruru.

      I haven’t had time yet. i’ll get to it.

    23. ruruland

      DRed: Sure, no problem.But don’t be a smug prick about other people not instantly replying.

      It’s not ab

      DRed: Sure, no problem.But don’t be a smug prick about other people not instantly replying.

      The difference is that Jowles and Owen almost nev

      DRed: Sure, no problem.But don’t be a smug prick about other people not instantly replying.

      It’s not about this time or any one time. It’s the consistent lack of responses and participation from one side.

    24. DRed

      ruruland: It’s not ab

      The difference is that Jowles and Owen almost nev

      It’s not about this time or any one time. It’s the consistent lack of responses and participation from one side.

      Owen literally spent years years on this site replying to people who question Berri’s metrics. The Berri people also have a website of their very own, and they respond to people who raise questions about their work. You’re more than welcome to go there and tell them you destroyed their model by showing the Nuggets had more rebounds with Chris Anderson than with Marcus Camby.

    25. Owen

      Frank –

      Look, I don’t want to sit here and defend every nuance of the Wages of Wins That’s not really how I care to spend my time. And frankly I am not enough of a zealot about the WOW (anymore) to care.

      The Wages of Wins came out during a very dark time for Knicks fans. It made the argument that scoring was overvalued, particularly inefficient scoring, during a time in which Knicks management was loading up on just that skillset. We all know how that worked out.

      It’s supposed to only be an objective starting point for analysis. I have been guilty in the (fairly distant past) of trying to end arguments by referencing WP. That’s really not what it’s for. It’s the start of the argument but not the finish. It’s an imperfect tool but much better than what existed previously, along with other measures.

      Re your questions-

      Defense – A significant portion of defense is captured in the box score stats, between rebounds, fouls, steals, team turnovers, etc. A significant portion is left out. The WOW answer, is that the element not captured, while not inconsequential, is a smaller component of total value than you would expect. Personally, i am ok with tacking on defensive data to any evaluation.

      Re – PF/C. You need different kinds of players. The model is built around certain assumptions about player roles, which are based off historical data. Faried does need someone to dribble the ball up the floor. Every big does. But that someone could be any one of 200 NBA quality guards in the player pool.

      Re Mo Williams – He was the same guy in Cleveland he was in Milwaukee, except he did get a bump up from their team defense. But he looked amazing playing next to Lebron. It’s a common thing. Players get overrated because they play next to stars.

      Re Fields – I said at the time it couldn’t last. The guy was a poor outside shooter in college. He shot better from behind the NBA arc as a rookie, then he had in college, and 75%+ from the line after…

    26. Owen

      ….after shooting 66% in college.

      One more thought. A season is not a huge sample. Especially at the beginning of a player’s career. Faried was over .300 last year. He is at .218 this year. It’s important to think carefully about sample size when looking at any data set. I certainly made that mistake David Lee’s rookie year when he was flat out amazing in relatively limited minutes. But you need more than one season of data to feel confident in any conclusion. The jury is still out on Faried, just like it would be on any player coming into the league.

      With a player like OJ Mayo, we have a pretty big sample. I am going to confidently predict (and yes I read the NBA Geek piece on this), that Mayo will not shoot closer to 40% than 50% this year from three, and probably will finish below 40%, while also seeing his ft% regress to 80%.

      You always have to take small samples with a grain of salt…..

    27. Juany8

      Let’s throw out a hypothetical situation. Say you have two teams with an overall TS% of 50 (to make it simple), and both use about 50 possessions every game on offense. Both have centers playing 36 minutes a game, and both centers get around 6 shots at the rim in those 36 minutes (including free throws). They both shoot 70 TS% on those shots, the only difference is that one center stops there, while the other has developed a jumper out to about 15 feet, and he takes around 6 shots a game away from the rim at a 54 TS% overall.

      Now from the example, it should be easy to see that one player ends up with a 70 TS% and the other player ends up with a 62 TS%, which is a pretty damn substantial difference. Was the first player actually more beneficial though? Both centers got the same number of shots at the rim at the same percentage, so they contributed exactly the same amount on those possessions. The other center also contributed on those other possessions though, at a higher percentage than the expected value of his teammates shots. Every shot he takes improves his team’s efficiency, so it’s good for him to take as many shots as possible.

      A shot’s value is largely dependent on the type of shots available at the given situation. For an extreme example, end of the quarter half court heaves are incredibly bad shots percentage wise, but always a smart play since it improves your possibility to scoring on that possession from 0 to 2%. So even though the second center has a noticeably lower TS% overall, he’s doing much more for his team’s offense by virtue of taking more shots away from the rest of the offense.

      The only argument I’ve heard to counter this is that the first center could simply increase his number of attempts overall, and that the distribution and efficiency of those shots will stay mostly the same. If that’s the case, why has Chandler’s usage never risen above 15 or Tim Duncan’s usage dropped below 20?

    28. Juany8

      By the way in the math example above, Player A (the Chandler type) needs his teammates to have a 47 TS% on the other 44 possessions the team has to shoot a 50 TS%. Player B (Tim Duncan type) only needs his team to shoot a 46 TS% on 38 other possessions. Now let’s say there’s a player C who doesn’t get ANY shots at the rim, and instead can only take jumpers at a 54 TS%. With the same math, his team would get to 50 TS% if he took 20 possessions at a 54 % rate, so that his teammates only need a 47 TS% on 30 shots to get to 50 TS%, and it’s easier to get the 47 for 30 possessions than for 44.

      Thus we see the value of Kobe and Melo, as long as the shots they’re taking are better than the average available shot in the situation, every shot they take is increasing their value while requiring less of the rest of the team to make the offense run.

    29. Frank

      Owen – thank you for a reasonable response. I’ll compose something later.

      THCJ – thank you for showing yourself yet again.

    30. DRed

      Juany, I think in your first example the jump shooting center would (all other things being equal) have more value according to WP. Even though he’s less efficient overall in terms of TS, he’s still making additional shots at the average efficiency for a center, so he’d be more valuable than the Chandler type player. Taking shots you hit 54% of the time is a good thing.

    31. ruruland

      ignoring the validity of the above statement, isn’t it still quite possible for a center to take slightly less than average efficiency shots if they are actually the best shots available.

      For example, you have Jason Kidd, Steve Novak, Pablo Prigioni, Tyson Chandler and Jr Smith on the floor.

      When the pick and roll gets stifled, the options are a)Jr Smith step-back (35%) b) a wild contested heave from Novak, Kidd or Prigioni (20%) or c) a Tyson Chandler post-up (45% and higher likelihood of beingfouled).

      I would say the above scenario with this group occurs quite frequently, most often resulting in JR Smith trying to create with his step-back.

      Say that happens 7 times a game. If the ball enters into Chandler all 7 times, that’s going to help the offense quite a bit while simultaneously destroying his efficiency.

    32. Juany8

      DRed:
      Juany, I think in your first example the jump shooting center would (all other things being equal) have more value according to WP.Even though he’s less efficient overall in terms of TS, he’s still making additional shots at the average efficiency for a center, so he’d be more valuable than the Chandler type player.Taking shots you hit 54% of the time is a good thing.

      DRed, you’re not supposed to focus so much on the actual numbers, really even a 50.5 TS% shot from a center would be valuable in the scenario I listed, he would just have to be able to take 40+ shots a game to get the same effect. It’s not about some arbitrary positional adjustment, it is by necessity an individual team adjustment. If the rest of your teammates are crap (think TD, Fields, Bibby, Walker…) and is shooting a 45 TS%, then you can shoot a 50 TS% and still be improving your team’s offense.

      Besides I’m not even remotely pointing out that guys like Duncan and Melo have a bigger impact on the possessions they’re not using with their passing and ability to space the floor.

    33. Juany8

      ruruland:
      ignoring the validity of the above statement, isn’t it still quite possible for a center to take slightly less than average efficiency shots if they are actually the best shots available.

      For example, you have Jason Kidd, Steve Novak, Pablo Prigioni, Tyson Chandler and Jr Smith on the floor.

      When the pick and roll gets stifled, the options are a)Jr Smith step-back (35%) b) a wild contested heave from Novak, Kidd or Prigioni (20%) or c) a Tyson Chandler post-up (45% and higher likelihood of beingfouled).

      I would say the above scenario with this group occurs quite frequently, most often resulting in JR Smith trying to create with his step-back.

      Say that happens 7 times a game. If the ball enters into Chandler all 7 times, that’s going to help the offense quite a bit while simultaneously destroying his efficiency.

      Exactly! As I said, no matter how inefficient a half court shot or end of the shot clock shot is, it’s still much better than letting the clock run out and getting no chance to score. WP would recommend a player never take a half court shot or late shot, the team gets the turnover instead of the individual. Maybe that scenario isn’t a big effect, but it breaks the basic assumptions behind a stat like WP, all shots are not created equal, shots can only be directly compared when the context is equal for both shots.

    34. DRed

      Juany8: Exactly! As I said, no matter how inefficient a half court shot or end of the shot clock shot is, it’s still much better than letting the clock run out and getting no chance to score. WP would recommend a player never take a half court shot or late shot, the team gets the turnover instead of the individual. Maybe that scenario isn’t a big effect, but it breaks the basic assumptions behind a stat like WP, all shots are not created equal, shots can only be directly compared when the context is equal for both shots.

      No it wouldn’t. Well, I mean technically it would lower your WP, or WS, or PER or whatever. But nobody would argue that you shouldn’t take half court shots at the buzzer because it hurts your WP. Juany, nobody is saying that WP tells you everything you need to know about a basketball player. I don’t know why you insist on thinking people who use it claim that it does, because none of us do.

    35. ruruland

      DRed: No it wouldn’t.Well, I mean technically it would lower your WP, or WS, or PER or whatever.But nobody would argue that you shouldn’t take half court shots at the buzzer because it hurts your WP.Juany, nobody is saying that WP tells you everything you need to know about a basketball player.I don’t know why you insist on thinking people who use it claim that it does, because none of us do.

      That’s pretty much exactly how it’s used. That’s why Jowles can show up when it’s really quiet and claim Ronnie Brewer is much more important to this team’s success than Carmelo Anthony, and why Owen believes that Tyson Chandler is better than Patrick Ewing was.

      Berri flacks boast about Marcus Camby being the best player in the NBA in 2007, when the Nuggets got considerably better without adding a player that was even half as valuable the next season.

      We can go on and on and on, but it’s sort of obvious that WP is not a very valuable metric outside of comparing rebounding role players, and even then it fails to account for the difference between grabbing uncontested rebounds and actually helping your team as a whole rebound.

    36. Owen

      yes, agreed DRed.

      JuanNY – It’s a long post basically setting up a strawman. WP would not recommend not taking a last second shot. That’s just silly.

    37. ruruland

      Owen:
      yes, agreed DRed.

      JuanNY – It’s a long post basically setting up a strawman. WP would not recommend not taking a last second shot. That’s just silly.

      What’s the strawman?

      He’s not just talking about last-second shots, that would be a strawman.

      He’s talking about all manner of shots created.

      Why can’t the Knicks increase Tyson Chandler’s usage above 15%?

      Is there any correlation between Chandler’s two highest usage years being among his two least efficient (19, 18 usage, .543, 565 usage)?

      Your post makes zero sense in relation to what we just posted. Surprise.

    38. DRed

      ruruland: That’s pretty much exactly how it’s used. That’s why Jowles can show up when it’s really quiet and claim Ronnie Brewer is much more important to this team’s success than Carmelo Anthony, and why Owen believes that Tyson Chandler is better than Patrick Ewing was.

      Berri flacks boast about Marcus Camby being the best player in the NBA in 2007, when the Nuggets got considerably better without adding a player that was even half as valuable the next season.

      We can go on and on and on, but it’s sort of obvious that WP is not a very valuable metric outside of comparing rebounding role players, and even then it fails to account for the difference between grabbing uncontested rebounds and actually helping your team as a whole rebound.

      Chris Paul had the highest WP in the NBA in 2007. His WP was much higher than Camby’s.

    39. DRed

      ruruland:

      Is there any correlation between Chandler’s two highest usage years being among his two least efficient (19, 18 usage, .543, 565 usage)?

      Yeah, maybe. You think there’s also a correlation between them being his age 19 and 20 seasons?

    40. ruruland

      DRed: Yeah, maybe.You think there’s also a correlation between them being his age 19 and 20 seasons?

      Paul Pierce, Garnett, Allen, right off the top.

    41. DRed

      ruruland: Paul Pierce, Garnett, Allen, right off the top.

      I have no idea what you think naming two guys who were in college when they were 19 and 20 has to do with Tyson Chandler. Do you not think that really young players tend to improve as they get older?

    42. ruruland

      DRed: I have no idea what you think naming two guys who were in college when they were 19 and 20 has to do with Tyson Chandler. Do younot think that really young players tend to improve as they get older?

      What?

      Those three guys had their most efficient seasons ear the end of the age curve when their usage went down from their primes.

      If Chandler’s efficiency wouldn’t change as his usage increases, why can’t the Knicks increase his usage?

    43. DRed

      jon abbey:
      Novak out tonight with the flu, we are running out of players.

      jon abbey:
      Novak out tonight with the flu, we are running out of players.

      Flu seems to be going around right now. Let’s hope nobody else gets it. Any word on Melo yet?

    44. jon abbey

      Melo is a gametime decision again, Sheed and Camby are both out again. Copeland looks like he’ll get serious minutes again, even if Melo plays (which I kind of doubt).

    45. Owen

      Ray Allen was great in Milwaukee. Great in Seattle, And great in Boston.

      Not really sure how his statistics are an argument against player performance being relatively constistent across different contexts.

      ruruland: What?

      Those three guys had their most efficient seasons ear the end of the age curve when their usage went down from their primes.

      If Chandler’s efficiency wouldn’t change as his usage increases, why can’t the Knicks increase his usage?

    46. ruruland

      jon abbey:
      looks like Melo is playing, hope that’s prudent but also yay.

      Don’t want that ankle to linger…. This has been a major issue throughout his career as he’s alluded to recently.

    47. ruruland

      Owen:
      Ray Allen was great in Milwaukee. Great in Seattle, And great in Boston.

      Not really sure how his statistics are an argument against player performance being relatively constistent across different contexts.

      Another strawman.

      The 6 most efficient seasons for Garnett were the 6th lowest in usage, four of which occurred after 31 when players are supposed to decline.

      Five of Pierce’s most efficient seasons came in five of his lowest usage seasons, four coming after the age of 30.

      Five of Allen’s most efficient seasons came in five of his lowest usage season. The only four times he eclipsed 600 TS was when his usage dipped below 21 and he was 33 years or older.

    48. ethsurken

      Is there any stat like (points per shot – league average pps) * number of shots? It seems easy to do and possibly useful.

    49. daJudge

      This may not be interesting or relevant, but it is kind of a response to the exclusive use of metrics in predicting future behavior or, as stated by CJ’s, explaining truth with a Capital T. I said this before and I will repeat it for comments. Metrics are used to predict future behavior in many contexts, often more serious than a b-ball game. For example, risk assessment tools are commonly used to predict recidivism for sexual offenders. I as well as others accept these tools as valid statistically, but not infallible. In fact, a jury really can not accept them as infallible. In my view, we properly assimilate and utilize common sense, experience, sub-conscious reasoning, logic and other faculties when we make judgments as to future behavior. To simply abstract a portion of an event and try to formulate a model to predict future behavior eliminates the above human characteristics and is therefore incomplete. Simply stated, such systems miss what is not measurable and often fail to properly measure factors contributing to particular effect. CJ, I think Bellow would concur that slavish adherence to a statistical formula is not a proper way to evaluate truth.

    50. Juany8

      daJudge:
      This may not be interesting or relevant, but it is kind of a response to the exclusive use of metrics in predicting future behavior or, as stated by CJ’s, explaining truth with a Capital T. I said this before and I will repeat it for comments.Metrics are used to predict future behavior in many contexts, often more serious than a b-ball game.For example, risk assessment tools are commonly used to predict recidivism for sexual offenders. I as well as others accept these tools as valid statistically, but not infallible.In fact, a jury really can not accept them as infallible.In my view, we properly assimilate and utilize common sense, experience, sub-conscious reasoning, logic and other faculties when we make judgments as to future behavior.To simply abstract a portion of an event and try to formulate a model to predict future behavior eliminates the above human characteristics and is therefore incomplete.Simply stated, such systems miss what is not measurable and often fail to properly measure factors contributing toparticular effect.CJ, I think Bellow would concur that slavish adherence to a statistical formula is not a proper way to evaluate truth.

      This is perfect. I also don’t see how it’s different from what me and Ruru are doing. I’ve shown plenty of studies, one by the guy who freaking runs WS, that shows players like Melo have value the box score doesn’t capture. What does it say when you point to Melo’s WS and the guy who runs it is explicitly telling you it is several wins too low. How can you then use that model to say Melo isn’t that good?

      Both me and Ruru have shown evidence of shot creation being important, often using legitimate statistical analysis. It’s not my fault THCJ and Owen instinctively dismiss any number that doesn’t fit with their chosen narrative

    51. ethsurken

      I think it’s obvious changing roles (and therefore types of shots, number of shots, style of play, etc) can effect efficiency and stats in general. Is anyone denying that?

    Comments are closed.