Statistical Analysis. Humor. Knicks.

Monday, October 20, 2014

Knicks Morning News (Thursday, Jan 26 2012)

  • [New York Times] Cavaliers 91, Knicks 81: Cavaliers Hand Knicks 7th Loss in 8 Games, 91-81 (Thu, 26 Jan 2012 06:20:11 GMT)
    Carmelo Anthony was 5 for 14 with 15 points as the Knicks committed 23 turnovers, which led to 24 Cavaliers points, and made only 42 percent of their shots.

  • [New York Times] N.B.A. Roundup: Deron Williams Leads Nets Past 76ers in Overtime (Thu, 26 Jan 2012 06:24:06 GMT)
    Deron Williams scored 34 points, including a go-ahead 3-pointer with 26.8 seconds left in overtime, to help lift the Nets to a 97-90 victory over the 76ers on Wednesday.

  • [New York Times] Lakers Get Last Word Over Clippers in Inner-City Clash (Thu, 26 Jan 2012 06:38:04 GMT)
    The inner-city battle for Los Angeles added another feisty chapter as the Lakers battled to a 96-91 win over the Clippers on Wednesday to reassume their top billing status.

  • [New York Times] Lakers Hold Off Clippers in City Rivalry Game (Thu, 26 Jan 2012 06:27:58 GMT)
    Kobe Bryant scored 12 of his 24 points in the fourth quarter, Pau Gasol had 23 points and 10 rebounds, and the Los Angeles Lakers rallied from a fourth-quarter deficit for a 96-91 victory over the Los Angeles Clippers on Wednesday night in another testy edition of their cross-hallway rivalry.

  • [New York Times] AP Source: Gordon Rejects Hornets’ Extension Offer (Thu, 26 Jan 2012 06:25:03 GMT)
    Guard Eric Gordon has turned down a four-year extension offer from the New Orleans Hornets, according to a person familiar with the situation.

  • [New York Times] Curry Has 32 Points, Warriors Top Blazers 101-93 (Thu, 26 Jan 2012 06:04:24 GMT)
    Stephen Curry scored a season-high 32 points and added seven assists and six rebounds, and the Golden State Warriors extended their recent home dominance against Portland with a 101-93 victory over the Trail Blazers on Wednesday night.

  • [New York Times] Gallinari Helps Nuggets Rout Kings (Thu, 26 Jan 2012 05:46:03 GMT)
    Danilo Gallinari celebrated a contract extension by scoring 21 of his 23 points in the first half and the Denver Nuggets routed the Sacramento Kings 122-93 on Wednesday night for their fifth straight victory and fifth in a row on the road.

  • [New York Times] Raptors Get First Win in Salt Lake City Since 2004 (Thu, 26 Jan 2012 05:40:11 GMT)
    Jose Calderon hit a 3-pointer to tie it at the end of regulation and the Toronto Raptors rallied from an 18-point first-quarter deficit to beat the Utah Jazz 111-106 in double-overtime Wednesday night.

  • [New York Times] Minnesota Wins 105-90 After Mavs Get Title Rings (Thu, 26 Jan 2012 04:13:18 GMT)
    Kevin Love had 31 points and 10 rebounds after signing his new contract and the Minnesota Timberwolves handed defending NBA champion Dallas another ceremonious loss, 105-90 Wednesday night.

  • [New York Times] Bonner, Blair Pace Spurs’ 105-83 Blowout of Hawks (Thu, 26 Jan 2012 04:16:11 GMT)
    Matt Bonner and DeJuan Blair both scored 17 points, and the San Antonio Spurs handed the surging Atlanta Hawks their first lopsided loss of the season, 105-83 on Wednesday night.

  • [New York Times] Granger, Hibbert Lead Pacers (Thu, 26 Jan 2012 03:52:02 GMT)
    Danny Granger scored 22 points Wednesday night and the Indiana Pacers became the first visiting team to win on Chicago’s home floor this season, beating the Bulls 95-90.

  • [New York Times] Thunder Top Hornets for 10th Win in 11 Games (Thu, 26 Jan 2012 03:55:18 GMT)
    Kevin Durant scored 25 points to lead the Oklahoma City Thunder to their 10th win in 11 games, 101-91 over the New Orleans Hornets on Wednesday night.

  • [New York Times] Jackson, Jennings Lead Bucks Over Rockets (Thu, 26 Jan 2012 03:37:07 GMT)
    Stephen Jackson and Brandon Jennings each scored 20 points, Ersan Ilyasova had a career-high 19 rebounds after Andrew Bogut sprained his left ankle, and the Milwaukee Bucks snapped an 11-game losing streak in Houston with a 105-99 victory over the Rockets on Wednesday night.

  • [New York Times] James’ Late FTs Give Heat 101-98 Win Over Pistons (Thu, 26 Jan 2012 03:37:00 GMT)
    LeBron James scored 32 points, including the game’s last six from the free throw line, to lead the Miami Heat to a 101-98 win over the Detroit Pistons on Wednesday night.

  • [New York Times] Off the Dribble: Wittman and the Wizards: Where to Begin? (Thu, 26 Jan 2012 02:27:13 GMT)
    The Washington Wizards are 2-15 and, not surprisingly, have a new coach, Randy Wittman.

  • [New York Times] Off the Dribble: Knicks Over Bobcats: That Was Supposed to Happen (Thu, 26 Jan 2012 02:26:13 GMT)
    The Knicks got a welcome victory over the Bobcats, but games against quality teams will be a better indication of where New York stands.

  • [New York Times] Off the Dribble: Hornets’ Gordon Offered Contract Extension (Thu, 26 Jan 2012 02:23:02 GMT)
    The Hornets have offered a contract extension to Eric Gordon, with the league’s approval.

  • [New York Times] Timberwolves, Love Reach 4-Year Extension (Thu, 26 Jan 2012 04:30:56 GMT)
    Kevin Love watched friends Derrick Rose and Russell Westbrook sign five-year extensions this season and was ready to do the same with the Minnesota Timberwolves.

  • [New York Times] Cavs Outhustle Knicks 91-81 (Thu, 26 Jan 2012 04:03:59 GMT)
    After he dressed following another disturbing loss, Carmelo Anthony was asked about his injured right thumb, the All-Star’s newest ailment.

  • [New York Times] Deron Williams, Humphries Lead Nets Past 76ers (Thu, 26 Jan 2012 03:04:09 GMT)
    Deron Williams scored 34 points, including a go-ahead 3-pointer with 26.8 seconds left in overtime, to help lift the New Jersey Nets to a 97-90 victory over the Philadelphia 76ers on Wednesday night.

  • [New York Times] Wizards Beat Bobcats 92-75 in Wittman’s First Game (Thu, 26 Jan 2012 02:43:04 GMT)
    Nick Young scored 20 points, and Andray Blatche had 17 points and 10 rebounds to lead the Washington Wizards to a 92-75 win over the Charlotte Bobcats on Wednesday night in Randy Wittman’s first game as coach.

  • [New York Newsday] Ailing Melo's only concern is his health (Thu, 26 Jan 2012 00:16:50 EST)
    The Knicks picked up next season's $2.068-million fourth-year option on point guard Toney Douglas' contract. The deadline to do it was Wednesday.

  • [New York Post] Curry slimmer, but still has beef with Knicks (Thu, 26 Jan 2012 04:14:05 -0500)
    DETROIT – A slimmer and rejuvenated Eddy Curry tried to keep his bitterness to a minimum as he anticipates tomorrow’s first showdown against the Knicks, but he admitted there are people he is not fond of in the organization.
    Curry, who had a terrible stint under Mike D’Antoni…

  • [New York Post] Douglas’ rookie contract option exercised (Thu, 26 Jan 2012 04:14:05 -0500)
    CLEVELAND â?? The Knicks made a commitment to combo guard Toney Douglas yesterday, exercising the final year option on his rookie contract, preventing him from becoming an unrestricted free agent this summer. Yesterday was the deadline.But Douglas didn’t repay them with a good showing last night in an awful…

  • [New York Post] Knicks lose to Cavaliers as Davis sits (Thu, 26 Jan 2012 04:06:44 -0500)
    CLEVELAND â?? The Knicks confirmed last night they will be a lousy team until point guard Baron Davis comes back.
    As currently constituted, this team stinks.
    Davis’ debut was not last night and it might not be tomorrow in Miami, where a South Beach slaughter may await. Coach Mike D’Antoni…

  • [New York Daily News] Knicks fall to Cavs for 7th loss in last 8 games (Thu, 26 Jan 2012 06:53:51 GMT)
    In their rehearsal for Friday’s showdown with the Miami Heat, the Knicks self-destructed again against a team with a losing record. Carmelo Anthony suffered through another awful shooting night and the Knicks were simply outworked by the Cavs, losing 91-81.

  • [New York Daily News] Knicks’ Davis still out, unlikely for Heat (Thu, 26 Jan 2012 05:54:18 GMT)
    Baron Davis didn’t make his Knicks debut against the team that released him Wednesday night, and it is unlikely that the point guard will be ready in time to face the Miami Heat on Friday.

  • 44 comments on “Knicks Morning News (Thursday, Jan 26 2012)

    1. njasdjdh

      After last night’s performances Gallo has jumped ahead of Melo in PER. So, not only are we losing the trade when you consider it as a whole, but we’re losing if you just limit it to Gallo and Melo. Of course, I don’t expect Melo to keep shooting THIS badly, but I also don’t expect his contract value to magically half either.

      I’m too lazy to check, but we’ll have cap space when LeBron’s Heat contract is up right?

    2. PaulStreetBoy

      njasdjdh:
      After last night’s performances Gallo has jumped ahead of Melo in PER. So, not only are we losing the trade when you consider it as a whole, but we’re losing if you just limit it to Gallo and Melo. Of course, I don’t expect Melo to keep shooting THIS badly, but I also don’t expect his contract value to magically half either.

      I’m too lazy to check, but we’ll have cap space when LeBron’s Heat contract is up right?

      i just wanted to re-post this comment as this is exactly the same as my feelings. Melo is a hell of a player, but then the trade that went through was very wrong and totally wiped out a first, or at most a second round playoff team.

      yet what do we have today? these Knicks are no better than that. really sad..

    3. The Honorable Cock Jowles

      Forget about PER; it’s nearly worthless. Have you seen Gallo’s WS48? Dude’s over .200. He’s been out of this world good. Playing next to Lawson and Nene doesn’t hurt.

    4. Roshi

      Can someone please summarize the major flaws with PER and WS/48? Seems that for some players there is a major discrepancy in how good they are according to these metrics.

      Also, is there any metric that summarizes defensive impact well? Steals/Rebs don’t tell much of the story…would be nice to know what a guy like Shump actually does to the player he’s guarding, as well as the whole opposing offense.

    5. Brian Cronin (@Brian_Cronin)

      The problem with PER is that it literally has a major mathematical flaw in it that the creator of PER has never adjusted for whatever reason.

      What he does is determine that every missed field goal costs a player 0.72 points. However, every made two-point field goal is worth 1.65 points and every made three-point field goal is worth 2.65 points. Therefore, if you work the math out, so long as a player hits 30% of his 2-point shots and something like 21% of his 3-point shots, the more the player shoots, the more the player gets credit towards his PER. The end result is that volume scorers are overrated, since you can be inefficient and still get credit if you take a lot more shots. There’s just no two ways about it – it is a major flaw that if he just adjusted it, PER would be a much better rating system.

      Also, do note that PER has never really been intended to measure defense. But since it doesn’t really purport to (even though it contains steals and blocks for some reason), then I don’t think that’s a “flaw,” per se.

      The major argument against WP/48 is just that in trying to make players’ productions match up with wins (as it is reverse engineered from wins), the creators incorrectly weight things like rebounds. They claim to have adjusted their formula this year. There’s no one “oh my god, that’s just totally wrong” mathematical flaw, but more of a general thing.

      The major argument against WS/48 is similar to the Wins Produced argument, which is that it is too much based on team success. The logic behind Win Shares is that each player had a share in wins, so if you work together what each player did, you should come up with the wins of the team. However, the argument is that on teams with a lot of success (like the Jordan-era Bulls) then there are extra “win shares” that have to be be accounted for, so role players on good teams get them and as a result, they might be over-rated. The counter is that the role players on the good teams just played better when surrounded by good teammates, but it is an interesting debate (you know, like did Ron Harper improve his play when all he had to do was play off the ball and be a tough defender as opposed to when he was asked to be a lead scorer or did Ron Harper just get a bump in Win Shares because he was playing with the greatest player of all-time?). Again, there’s no “oh man, that is just a glaring error!” type flaw like there is with PER, just a general “role players appear to improve when playing on good teams.” The stars generally stay the same (like Dwyane Wade and Chris Bosh saw no increase to their Win Shares when they began playing with Lebron, Tyson Chandler’s Win Shares stayed the same going from the World Champion Mavericks to the…you know…Knicks, etc.).

      Win Produced tends to try to get exactly matched up with team wins, while Win Shares tends to leave a little more room for luck in determining team wins (you know, like, “if you add together all the Win Shares on a team and you get 50, there’s no guarantee you’ll actually get a 50 win team – you just have a team that should be good enough to win 50 games” while Wins Produced will almost always add up to very close to the actual win total of the team, typically within 10% of the actual win total – and it is that desire to be sooo close to the actual win total that leads some to think that some players’ Wins Produced stats are skewed a bit). As an example, the Knicks currently have a little under 9 win shares, while they have only 7 actual wins. Therefore, Win Shares theorizes that the Knicks have a team that should be good enough to be 9-9 rather than 7-11. Which does seem to make sense.

    6. Bruno Almeida

      from what I know, PER really rewards high volume scorers, even if their efficiency is not good… Kobe Bryant, for example, is 4th on PER this season, but only 39th in WS/48…

      but others can definitely explain the differences better than me.

    7. gransoporro

      The Honorable Cock Jowles: Forget about PER; it’s nearly worthless. Have you seen Gallo’s WS48? Dude’s over .200. He’s been out of this world good. Playing next to Lawson and Nene doesn’t hurt.

      During Summer and the lockout he played for Italy and Milan but he didn’t set the house on fire. The 3-4 times I saw him was his usual NY self at 14PPG. He is much more improved now, especially in terms of intensity.

      And of course all of the above is based on my observation: don’t ask me for statistics to back it up (anyway the size and heterogeneity of the sample would make any statistics useless…)

    8. A Voice of Reason

      Call me old fashioned, but the eye test really works… Granted you have to watch many games but I love the NBA and I don’t mind. When an educated fan watches a game, they can see who is responsible for wins. I can tell who rotated on defense to prevent an easy entry pass to a post player with position. I can also see who fell asleep and let a six foot guard get an uncontested layup along the baseline. God bless 1080i

    9. Roshi

      Thank you, very helpful. I am always intrigued by how rebounding is counted in these formulas. Observationally, some rebounds seem to be worth a lot more than others. An offensive rebound in traffic which leads to a score is more significant than a defensive rebound with no one else around, no? Granted, watching the games is not definitive proof, but it appears that Tyson Chandler’s rebounding generates a ton of extra possessions and points for us indirectly, whereas Bill Walker, for example, just grabs the rebounds that are readily available to him.

    10. The Honorable Cock Jowles

      The difficulty with wins is that they’re of a binary sort. A team could lose ten games by one point each and have the same record as a team that loses by thirty. Of course, pythagorean wins (which are the best way I know to determine a team’s actual value) suggests that a team with a -1.0 point differential should have something like a 40-42 record by season’s end, but with play-through injuries, bad ref calls, and freak occurrences (80-foot buzzer beaters, etc.) there’s always room for error.

    11. Brian Cronin (@Brian_Cronin)

      Thank you, very helpful. I am always intrigued by how rebounding is counted in these formulas. Observationally, some rebounds seem to be worth a lot more than others. An offensive rebound in traffic which leads to a score is more significant than a defensive rebound with no one else around, no? Granted, watching the games is not definitive proof, but it appears that Tyson Chandler’s rebounding generates a ton of extra possessions and points for us indirectly, whereas Bill Walker, for example, just grabs the rebounds that are readily available to him.

      Offensive rebounds are deemed more valuable than defensive rebounds in all of the various metrics.

    12. Roshi

      By the way, I remember someone here last season jokingly proposing a “Catching the Ball Situations Sucess %” stat last year to quantify Mozgoz’s stone hands. I wish someone tracked ‘Blow Byes Surrendered’. Amare might be last in the league, followed closely by Toney Douglas. Re Amare, how can someone have such a quick first step on offensive, yet have no lateral quickness on defense?

    13. Brian Cronin (@Brian_Cronin)

      Call me old fashioned, but the eye test really works… Granted you have to watch many games but I love the NBA and I don’t mind. When an educated fan watches a game, they can see who is responsible for wins. I can tell who rotated on defense to prevent an easy entry pass to a post player with position. I can also see who fell asleep and let a six foot guard get an uncontested layup along the baseline. God bless 1080i

      The thing is, there is a false dichotomy set up between stats and “the eye test.” The two are not opposing beliefs. In-person scouting is by far the most valuable talent evaluator in the NBA. No doubt about it. Stats just work as a way to support in-person scouting, not to supplant it. For instance, a good chunk of us here watch every Knick game (or at least we did when Time Warner Cable aired Knick games), so we can more or less see what the stats will tell us without looking at the stats. We can all see that Melo is highly inefficient in isolations. We all can see that Amar’e is even worse in isolations. However, not everyone is watching the game, which is where stats come in. Especially when it comes to teams you aren’t watching every game of. Stats support in-person scouting. It is the way every basketball front office does it – the only difference is that while every team highly values in-person scouting, some teams don’t value advanced stats as much. So the difference between teams is not how they treat in-person scouting, it is just how they treat advanced stats. It is the same thing in baseball – the biggest “stat” teams, like the Oakland Athletics, for instance, are heavily geared around in-person scouting. They just also use advanced stats to support their in-person scouting. The difference is that some teams don’t use the advanced stats very much at all.

    14. Brian Cronin (@Brian_Cronin)

      The difficulty with wins is that they’re of a binary sort. A team could lose ten games by one point each and have the same record as a team that loses by thirty. Of course, pythagorean wins (which are the best way I know to determine a team’s actual value) suggests that a team with a -1.0 point differential should have something like a 40-42 record by season’s end, but with play-through injuries, bad ref calls, and freak occurrences (80-foot buzzer beaters, etc.) there’s always room for error.

      Oh definitely, pythagorean wins are clearly the best way to determine how good a team is.

    15. The Honorable Cock Jowles

      But there’s also a problem with using stats to support the eye-test, and that’s the deficiency of human memory. It’s important to not discredit Kevin Martin as a shooter just because he’s not as athletic as John Wall or Iman Shumpert.

      It would be impossible for anyone (who’s not a compulsive counter) to tell the difference between a 50 TS% shooter and a 60 TS% shooter with the eye-test and memory alone. Yet that difference between the two could be the difference between a top 10 player and a top 300 player. And as we know, you pretty much need a top 10 player to make the Finals. So I actually advocate the opposite: find a statistical model that works, and then use subjective evaluation to determine why certain players are effective playing next to other players, and draft, sign, or trade players with those specific skillsets to fill your team’s needs.

    16. Richmond County

      I’m out on the West Coast and frequently have business meetings during games played in the East so I’m stuck following games using Gamecast on my phone. Reading line after line of “T. Douglas misses from 20″ and “A. Stoudemire misses from 18″ is down right painful. You can literally read the lack of ball movement and offensive cohesion through the one-line descriptions of shot selection.

    17. Scorpio Dragon

      A lot of defense is preparation and anticipation more than pure athletic ability. It’s why players like Bowen and Battier were/are great defenders, yet athletically were/are considered sub-par and limited in the NBA. If you truly study a player’s tendencies under a myriad of conditions, you should be able to defend them relatively well. Part of the reason A’mare turns the ball over so much via bad pass or offensive foul is that he’s become so one dimensional that the defender can anticipate what he’s going to do.

      I was doing some digging on 82games and found that the difference between our #1 line up and #2 line up is significant. Not so much on the Offensive end as it is on the Defensive end.
      Lineup Mins OffRtg DefRtg
      1 Shumpert-Fields-Anthony-Stoudemire-Chandler 96 1.01 .90 2 Douglas-Fields-Anthony-Stoudemire-Chandler 70 1.05 1.09

      Sure it’s a small sample size but if it stays consistent and Davis can be an average point guard, I think the team might just be alright.

    18. Scorpio Dragon

      That should read
      Lineup Mins OffRtg DefRtg
      1 Shumpert-Fields-Anthony-Stoudemire-Chandler 96 1.01 .90
      2 Douglas-Fields-Anthony-Stoudemire-Chandler 70 1.05 1.09

      Scorpio Dragon: Lineup Mins OffRtg DefRtg
      1 Shumpert-Fields-Anthony-Stoudemire-Chandler 96 1.01 .90 2 Douglas-Fields-Anthony-Stoudemire-Chandler 70 1.05 1.09

    19. Juany8

      @15, that’s not how science works, you develop a hypothesis based on observation and then attempt to test and prove your theories, often, but not always, with math. If by “find a statistical model that works” you mean use a math formula to determine player value then using fit player evaluations to the model, you’re literally doing the opposite.

      For instance, it’s been pretty widely accepted that having a high eFG%, high rebounding (especially offensive boards), and high FTA, coupled with low turnovers, and forcing your opponent to do the opposite, correlates incredibly well to winning through various statistical model. Those numbers, however, did not just evolve on their own through statistical analysis. Every single coach in the league advocates for taking high percentage shots, limiting turnovers, getting to the free throw line, and crashing the boards. From that knowledge, it’s not highly surprising that the four factors correlate statistically well with winning, basic observation would suggest as much, and further analysis gives the results you see today.

      The next step would be to really attempt to quantify what causes the four factors at a team level. As someone said earlier, Chandler battles and boxes out for a lot of tough rebounds, whereas someone like Douglas mostly catches balls that bounce to his general area. Chandler therefore helps the team get rebounds at a general level because of his boxing out and activity, and high value rebounds that would otherwise become offensive rebounds at point blank range for the other team. Similarly with high efficiency shots, every Knicks fan would love if Chandler could take the number of shots Amar’e and Melo take, and no coach fails to notice that Chandler almost never misses, but it simply takes a lot of offensive effort to get Chandler the shots he can take efficiently. That extra level of “effort” (be it a pass, pick and roll, etc.) needs to be quantified before making any conclusions.

    20. Roshi

      Juany8: but it simply takes a lot of offensive effort to get Chandler the shots he can take efficiently.

      I think this is a point that some stats-oriented people on this site miss. Yes, Chandler’s TS% is through the rough, but that doesn’t mean he can maintain that increasing his FG attempts to 25 a game. Chandler’s buckets come almost entirely on put-backs and pick-and-roll dunks. Even the best offenses don’t have those kind of opportunities every time down the floor. Granted, he’s one of the best finishers in the league on those plays, but we’d definitely see a substantial drop in his TS% if we asked him to be a focal point of the offense, as his shot creation and back to the basket skills are pretty mediocre. I’d love to be able to run the pick-and-roll with him more often, but with our terrible outside shooting, I would expect defenders to shadow Chandler (or Amare) and live with our guards taking the open 15 footer.

    21. Brian Cronin (@Brian_Cronin)

      But there’s also a problem with using stats to support the eye-test, and that’s the deficiency of human memory. It’s important to not discredit Kevin Martin as a shooter just because he’s not as athletic as John Wall or Iman Shumpert.

      It would be impossible for anyone (who’s not a compulsive counter) to tell the difference between a 50 TS% shooter and a 60 TS% shooter with the eye-test and memory alone. Yet that difference between the two could be the difference between a top 10 player and a top 300 player. And as we know, you pretty much need a top 10 player to make the Finals. So I actually advocate the opposite: find a statistical model that works, and then use subjective evaluation to determine why certain players are effective playing next to other players, and draft, sign, or trade players with those specific skillsets to fill your team’s needs.

      I dunno, doesn’t that still go with the idea of the stats supporting the eye test? I absolutely agree that the naked eye can’t tell if a player is a 55% TS or a 60% TS (I think they actually could with a 50 and 60 TS%), just like the naked eye can’t tell the difference between a .250 hitter in baseball versus a .300 hitter. But the naked eye can tell you that Kevin Martin is a very good shooter. You just watch him in a game and he hits a lot of his shots and he shoots from the outside a lot, so your naked eye would tell you that this Martin guy seems to be a really good shooter. Then you check the stats and you see just how good.

      Stats definitely come into play is to keep a scout from being too obsessed with “athletic” bodies. I totally agree that “athletic” players get overrated and that’s where stats help you put things into perspective. That was one of the major points of Moneyball – that stats keep teams from drafting players too much on how the players look. However, again, you wouldn’t have found the player in the first place if it were not for scouting. DeJuan Blair, for instance, would never have become a college star just by someone looking at his high school stats. He had to be scouted in person.

      Also, “naked eye” does have to be “unbiased naked eye.” I agree that some people will look at a player score a lot and say “That guy is a great scorer!” and overrate him (see Jamal Crawford). However, an unbiased naked eye will see that Crawford missed a whole lot of shots while getting his points. The stats will then confirm that. So yes, biases have gotten in the way of the naked eye for decades in the NBA. Even then, though, wasn’t it sort of a misunderstanding of stats, also? Points per game is a stat, ya know? The NBA just overrated its importance for decades (heck, it still does).

    22. Brian Cronin (@Brian_Cronin)

      I think this is a point that some stats-oriented people on this site miss. Yes, Chandler’s TS% is through the roof, but that doesn’t mean he can maintain that increasing his FG attempts to 25 a game. Chandler’s buckets come almost entirely on put-backs and pick-and-roll dunks. Even the best offenses don’t have those kind of opportunities every time down the floor. Granted, he’s one of the best finishers in the league on those plays, but we’d definitely see a substantial drop in his TS% if we asked him to be a focal point of the offense, as his shot creation and back to the basket skills are pretty mediocre. I’d love to be able to run the pick-and-roll with him more often, but with our terrible outside shooting, I would expect defenders to shadow Chandler (or Amare) and live with our guards taking the open 15 footer.

      I don’t think any Chandler fan would believe that increasing his usage wouldn’t decrease his efficiency. We know that increases in usage typically correlate with decreases in efficiency. Chandler’s TS% are sooo high, though, that he can decrease his efficiency a whole ton and still be a very efficient player.

    23. Roshi

      On that point, why is it that after a decade in the league, Chandler hasn’t learned any legitimate back to the basket moves? Surely, he has time in practice and summer to perfect a sweeping hook. Would be nice to see him add this to his game, instead of just being our garbage points and dive man for the next five years.

      @25, having said that, from a coaching standpoint how do you identify which low usage, low mpg guys are ripe to maintain their high shooting efficiencies with increased minutes or offensive responsibilities? Seems to have worked out for Millsap, but can’t see it happening for Chandler.

    24. bobneptune

      Bruno Almeida:
      from what I know, PER really rewards high volume scorers, even if their efficiency is not good… Kobe Bryant, for example, is 4th on PER this season, but only 39th in WS/48…

      but others can definitely explain the differences better than me.

      so which do you think is closer to the truth….. is tyson chandler 32 players closer to the top than kobe?

      just asking.

    25. Juany8

      Well part of the problem with just looking at overall efficiency is that it doesn’t really tell you what you can do on a given game. Yes it would be ok if he got a few more shots and his efficiency decreased a bit, but not if it resulted in Chandler jumpers, post ups, or drives, all which would be a horrendous offensive result. I know it would be hard, but something like DVOA from Football Outsiders, where the result of the play is compared to the expected result for every single play of the game to get a ranking of players and teams. A point guard can just start taking more shots and driving more (as can Melo and Stoudemire). The result of the play won’t be great, but it will work often enough to not sink your team. The defense is capable of entirely dictating how often Chandler shoots by just staying between him and the rim in a half court offense.

    26. Z

      The Honorable Cock Jowles:
      with play-through injuries, bad ref calls, and freak occurrences (80-foot buzzer beaters, etc.) there’s always room for error.

      Isn’t this a major problem with basketball stat keeping? The fact that they don’t account for subjectivity, of which there is a lot in basketball? Not only do the refs have a huge impact on the outcomes of games, but they have a huge outcome on individual stats, too. A foul can be called on every possession. It is up to the refs to NOT call them, for the sake of continuity and crowd entertainment.

      Plus, we know that refs aren’t even basing their calls on their own subjectivity all the time, but on others’ as well: the commissioner’s, the media’s, the mafia’s, etc… Whether it is to keep games close, or to market superstars, or to keep the fans in Detroit from rioting, it seems that the refs have as much to say about win shares than the actual players do.

      (And then there is the subjectivity of the official scorer, as jon has pointed out a lot over the years.)

      So much of the stat keeping process is arbitrary, I don’t think a real scientist would find the numbers very useful.

    27. Roshi

      I may be in the minority, but I actually think that NBA refs do an exceptional job getting the calls right 95+% of the time, and at least compared to football and hockey refs, they are a revelation.

      And no, this is not David Stern posting!

    28. Z

      Z:
      So much of the stat keeping process is arbitrary, I don’t think a real scientist would find the numbers very useful.

      I should add that some advanced stats are obviously a zillion times better than the dinosaur stats that the mainstream still uses. Per game stats, this site was quick to convince me, are useless, and I can’t fathom why TS% hasn’t replaced FG% completely.

      It’s just the blind adherence to “the numbers say it, and numbers don’t lie” that I continue to reject. In baseball, which can be umpired by a computer, perhaps, but not basketball.

    29. The Honorable Cock Jowles

      Brian Cronin (@Brian_Cronin): I dunno, doesn’t that still go with the idea of the stats supporting the eye test?

      I think my point is that we shouldn’t be altering our models because they don’t match the eye-test. Here’s a great post from Wages of Wins that illustrates my point:

      http://wagesofwins.com/2009/03/29/there-is-not-much-difference-between-danny-granger-and-kobe-bryant/

      “And I would add, the lack of formal training also hampers the ability of sports writers to evaluate statistical models. Frequently the results of a model are simply matched to what the sports writer previously believed. If there is a match (as we see with something like PERs), then the model is “good.” If there isn’t a match (as often happens with Wins Produced, Adjusted Plus-Minus, or Dean Oliver’s work) then the sports writer calls for more research.

      This approach to research, though, is incorrect. Research is not done to confirm what we already believe. Research is done to teach us something new. Of course that doesn’t mean that all “new research” is “good.” In fact, much of what I see on-line with respect to sports would never be considered “good” by a peer-reviewed academic journal.

      But we don’t reach the conclusion of “good” or “bad” by checking the results of the research against our prior beliefs. No, “good” or “bad” is determined by looking at how the research was done and interpreted.”

    30. Z

      I read the Eddy Curry story from the morning news links. “Curry Has Beef With Knicks”. I clicked on it, irate at the notion that Eddy Curry can be mad at us for anything.

      Turns out he’s not. It’s a total bullshit headline. Curry says in the article: “I’m just looking toward the future right now. I don’t feel any way, positive or negative. I don’t feel anything.”

      Like FG%, and per game stats, I can’t fathom why anybody would read the mainstream media when sites like KB exist.

    31. The Honorable Cock Jowles

      Z:

      It’s just the blind adherence to “the numbers say it, and numbers don’t lie” that I continue to reject.

      In my own defense, I don’t believe that the numbers are 100% accurate. I just think they’re much closer than any subjective test could be. I am aware that Chandler has a limited offensive game, but considering that there are and have been other centers in the league who play or played with the same, in-the-paint, no-jump-shot game that he does, yet he is able to produce with an efficiency that has never been seen before in this league, I believe that he’s a top 3 center, behind Howard and Bynum. I have some stats to back that up, but hell, they could be wrong. I just think they’re less likely to be wrong than the argument that because he can’t do the Dream shake (etc.), he’s not an immensely effective player who is worth every penny of his contract (unlike our two max “stars” who are responsible, directly and/or indirectly, for the sub-mediocrity we’ve experienced).

      I just don’t think that I’ve been wrong very often since I started reading Dave Berri. DeJuan Blair, Ty Lawson, Fields, Faried (you have to admit that he’s been terribly impressive in his limited minutes, preseason on), Carmelo, Amar’e, et al. Small smaple, maybe, but his numbers have worked out pretty well for what would be a tremendously efficient young team on rookie contracts.

    32. bobneptune

      Robert Silverman (@BobSaietta):
      S’okay guys. All will be fine once Dolan hires me as coach: http://knickerblogger.net/an-open-letter-to-james-dolan/

      tell him he needs to sign a sane person as owner!

      walsh , for all his flaws , had the team moving in the right direction with a ton of flexibility and options moving forward til little jimmy d wanted to raise his ticket prices through the roof and he needed a “star” to sell to all the pseudo world’s greatest fans.

      it jsut feels so special being played as a mark.

    33. Brian Cronin (@Brian_Cronin)

      I think my point is that we shouldn’t be altering our models because they don’t match the eye-test. Here’s a great post from Wages of Wins that illustrates my point:

      http://wagesofwins.com/2009/03/29/there-is-not-much-difference-between-danny-granger-and-kobe-bryant/

      “And I would add, the lack of formal training also hampers the ability of sports writers to evaluate statistical models. Frequently the results of a model are simply matched to what the sports writer previously believed. If there is a match (as we see with something like PERs), then the model is “good.” If there isn’t a match (as often happens with Wins Produced, Adjusted Plus-Minus, or Dean Oliver’s work) then the sports writer calls for more research.

      This approach to research, though, is incorrect. Research is not done to confirm what we already believe. Research is done to teach us something new. Of course that doesn’t mean that all “new research” is “good.” In fact, much of what I see on-line with respect to sports would never be considered “good” by a peer-reviewed academic journal.

      But we don’t reach the conclusion of “good” or “bad” by checking the results of the research against our prior beliefs. No, “good” or “bad” is determined by looking at how the research was done and interpreted.”

      Oh sure, I agree with that. I am more coming from a position that the good players can be seen by the naked eye and that stats just support it and help you compare players that you don’t see in person. I’ve yet to see a player whose Win Shares are great that didn’t look great in person when watching them play regularly. Take Chandler, for instance, stats say he’s awesome and he looks awesome out there.

    34. Juany8

      Again THCJ, you totally misunderstand how science works, you develop a prior belief (hypothesis) through observation and then you attempt to test it scientifically. Testing it scientifically does not mean holding it up against a pre-conceived formula, because you’re just testing a hypothesis at this point, you can’t already have a conclusion you’re testing it against. WP and WS and PER and +/- are all based on assumptions about how basketball works, and what the intrinsic value of certain statistics are (who says an average or a linear correlation have inherent meaning? They’re just numbers). Essentially they are conclusions made without doing research. By holding up your observations (watching the games, which is considered scientific data) to a formula of how someone thought player value should be computed, you’re doing the same kind of bad research you accuse others of doing. There is no inherent meaning in stats beyond which can be physically contextualized, numbers in a vacuum don’t tell anyone anything

    35. JK47

      “I think my point is that we shouldn’t be altering our models because they don’t match the eye-test.”

      Well, yes and no.

      The second Bill James Historical Baseball Abstract was the first place I ever remember seeing the Win Shares theory. James’ version of Win Shares, at that time, had advanced statistics for fielding, accounting for things like groundball/flyball ratios, amount of left-handed vs. right-handed batters a defense faced… things like that. Sometimes the system seemed to work very well– players who were obviously underrated by previous, more primitive fielding systems like Linear Weights fared much better in the more advanced Win Shares.

      But Win Shares also produced some rather iffy results. James used Win Shares to create “Hypothetical Gold Gloves,” in which he retroactively assigned Gold Gloves to players who had led their position in defensive Win Shares that season. Some of the results were embarrassing. I don’t have the book here in my office but there were some truly awful defensive players who won “Win Shares Gold Gloves.”

      James admitted this, essentially admitting that at this point in their development, Win Shares sometime failed the “eye test” and needed to be tinkered with further. It’s common sense: if your statistical model has Dick Stuart or Lonnie Smith winning Gold Gloves, you know it needs work. Same thing for Wins Produced: Landry Fields supposedly ranking as the 12th most productive player in the NBA is a broken result, something that reveals that the system needs to really improve. Call it the “eye test” if you want, but that is a ludicrous result. Landry Fields is an extremely limited player, not an elite player. Period. Yet the Wins Produced gang doesn’t seem to admit this, they insist he was one of the best players in the league last year.

      Needs more eye test.

    36. Z

      Roshi:
      I may be in the minority, but I actually think that NBA refs do an exceptional job getting the calls right 95+% of the time…

      Wow, and that’s after watching Anderson Varejao flop his way up and down a basketball court for 40 minutes…

    37. bobneptune

      Z: Wow, and that’s after watching Anderson Varejao flop his way up and down a basketball court for 40 minutes…

      did he flop his way to 14 boards ad 2 blocks last night? he out hustled and out played the statistically magnificent one last night.

    38. Bruno Almeida

      bobneptune: so which do you think is closer to the truth….. is tyson chandler 32 players closer to the top than kobe?

      just asking.

      Look, I’m not saying the advanced stats are 100% accurate all the time, but an argument can certainly be made that Kobe has been widely overrated during his whole career.

      I know, he’s got 5 titles, but 3 of those came playing second-fiddle to the clear cut best big man of his era (and one of the top 15 players ever) and the other 2 with a number of really good, really efficient players in Pau Gasol, Lamar Odom and others.

      In this particular season, with Kobe chucking it like it’s 2005, I’d definitely argue that Chandler has a bigger positive impact on his team than Kobe has on his… Kobe has a career high usage rate on a career low TS%.

    39. hoolahoop

      The Honorable Cock Jowles:
      But there’s also a problem with using stats to support the eye-test, and that’s the deficiency of human memory. It’s important to not discredit Kevin Martin as a shooter just because he’s not as athletic as John Wall or Iman Shumpert.

      It would be impossible for anyone (who’s not a compulsive counter) to tell the difference between a 50 TS% shooter and a 60 TS% shooter with the eye-test and memory alone. Yet that difference between the two could be the difference between a top 10 player and a top 300 player. And as we know, you pretty much need a top 10 player to make the Finals. So I actually advocate the opposite: find a statistical model that works, and then use subjective evaluation to determine why certain players are effective playing next to other players, and draft, sign, or trade players with those specific skillsets to fill your team’s needs.

      However statistics miss soooo much that they eye can see. Yes, a valuable tool, to a degree, stats alone simply cannot begin to capture the true picture of a player.
      You cannot put an accurate number value on players based on the current metrics, without making a lot of errors.

    Comments are closed.