Statistical Analysis. Humor. Knicks.

Monday, November 24, 2014

Knicks Morning News (Friday, Oct 05 2012)

  • [New York Daily News] Knicks’ Kurt Thomas still fit at 40 (Fri, 05 Oct 2012 03:29:00 GMT)
    The body is fine and the mind is still sharp.

  • 32 comments on “Knicks Morning News (Friday, Oct 05 2012)

    1. Juany8

      Lol Shaq might be the dumbest analyst of all time. He killed the TNT show whenever he talked, he just doesn’t have a good understanding of basketball, or he’s just trying to make a dumb point for publicity. Now I don’t think Brook Lopez is the worst center in the league, but I do think he might have the worst contract in the NBA, I don’t know that I’d give him even as much as McGee got. He’s just crap at defense and can’t rebound, and his starting front court mate can only rebound but not defend. If the Nets were in the West, they wouldn’t sniff the playoffs

    2. thenamestsam

      Shaq is just trolling. He may be a moron, but you could take a 5-year old to their first NBA game and they could see that Dwight is better than Lopez.

      Shaq just loves the attention, and there’s about ten ways he could get his name in the news at this point, and the other nine of them start with the phrase “Shaq was arrested today for…” So it’s not hard to see why he went with this option. He’s irrelevant, and he can’t stand it. I wouldn’t be shocked if we started hearing rumblings about an attempted comeback sometime in the near future.

    3. d-mar

      It’s all a part of the endless Brooklyn Nets lovefest going on with the media (concurrent with the trashing of the Knicks.)

      Only way to stop it is to kick the Nets asses on opening night in their new building.

    4. The Honorable Cock Jowles

      ruruland: If steals were the most important aspect of defense you would see a lot more of them. There have been a lot of great thieves who were poor defenders. One of them is mentioned in this thread.

      To say steals = net possessions on defense … no words for how ridiculous that is.

      I’d like to see evidence that suggests that high steal percentage correlates with low opponent eFG%.

    5. iserp

      THCJ, be a little more scientific. Elaborate, if what ruruland says is true, why would it correlate looking at actual data points?

    6. chrisk06811

      I think the ‘sheed signing was so that we don’t have to fight about stats I don’t have time to understand anymore.

      ball don’t lie.

    7. The Honorable Cock Jowles

      iserp:
      THCJ, be a little more scientific. Elaborate, if what ruruland says is true, why would it correlate looking at actual data points?

      He seems to imply that steals affect defensive efficiency due to aggressiveness and “risk-taking.” He says (anecdotally) that Camby is a great shot-blocker and thus a poor positional defender, so I assume the same is true about steals and positional play.

      I just don’t see a correlation between defensive TOV% and defensive eFG%.

    8. The Honorable Cock Jowles

      chrisk06811:
      I think the ‘sheed signing was so that we don’t have to fight about stats I don’t have time to understand anymore.

      ball don’t lie.

      His ball may not lie, but neither does the L column.

    9. iserp

      The Honorable Cock Jowles: He seems to imply that steals affect defensive efficiency due to aggressiveness and “risk-taking.” He says (anecdotally) that Camby is a great shot-blocker and thus a poor positional defender, so I assume the same is true about steals and positional play.
      I just don’t see a correlation between defensive TOV% and defensive eFG%.

      As usual you don’t answer what i ask, but what you want to say.

      I only asked because even if ruruland is right, you might not see the correlation
      you “expect”. That could be because both observables are correlated to begin with. I’ll make some hypothesis to illustrate: Players are parametrized by “arm length” and “aggressiveness”, and i measure them with two observables “%eFG allowed” and “%STL”. “aggressiveness” increases both %eFG and %STL, while “arm lenght” increases %STL but decreases %eFG allowed . If i look at the correlation between %STL and %eFG, it could be positive, negative or uncorrelated, just depending in the strength of the contributions of the parameters to the observables.

      So again, what correlation would you expect between %STL and %eFG allowed if the non quantifiable and unscientific parameter “aggressiveness” didn’t play any role at all?

      When you answer that question, we can look at real data and compare. If the correlation is higher than expected blah blah blah, if it is lower than expected blah blah blah, if it is the expected, “aggressiveness” doesn’t have any influence on %eFG allowed.

    10. Frank

      re: this whole stl% and efg% argument –

      there’s way too much noise.
      Some guys have a good stl% because they just have very quick hands and good anticipation while staying in position like Shump or Tony Allen. Some guys might have a good stl% because they gamble like crazy, leaving themselves out of position and their teammates have to help, causing rotations until the defense breaks, like JR Smith. Trouble is, stl% does not distinguish between these two. eFG-allowed is confounded by the overall team defense.

      For lack of having good stats to measure how much someone gambles, how much someone doesn’t gamble, I think we can just watch tape of guys like JR or Toney Douglas (just to name 2 guys we see all the time), and see that they gamble too much and get blown by because of it.

    11. ruruland

      Frank:
      re: this whole stl% and efg% argument –

      there’s way too much noise.
      Some guys have a good stl% because they just have very quick hands and good anticipation while staying in position like Shump or Tony Allen. Some guys might have a good stl% because they gamble like crazy, leaving themselves out of position and their teammates have to help, causing rotations until the defense breaks, like JR Smith.Trouble is, stl% does not distinguish between these two.eFG-allowed is confounded by the overall team defense.

      For lack of having good stats to measure how much someone gambles, how much someone doesn’t gamble, I think we can just watch tape of guys like JR or Toney Douglas (just to name 2 guys we see all the time), and see that they gamble too much and get blown by because of it.

      of course. Stl % means very little. To use it as a marker of overall efficiency and player value is absurd.

    12. njasdjdh

      Jesus Christ, ruru…please stop. I’m fairly certain that wasn’t his point and I’m pretty sure you know that. The poster mentioned STLs and TOs and was trying to point out how awesome it is, given that we know that possessions matter, that Chris Paul is so good at a.) not turning the ball over and b.) ending the opposition’s possession via STL that he damn near averages more STLs than TOs, which is almost unfathomable. Stop trying to make it more than it is in your attempt to belittle any poster who dares present statistically bent posts that don’t comply with your world view.

      ruruland: of course. Stl % means very little. To use it as a marker of overall efficiency and player value is absurd.

    13. ruruland

      njasdjdh:
      Jesus Christ, ruru…please stop. I’m fairly certain that wasn’t his point and I’m pretty sure you know that. The poster mentioned STLs and TOs and was trying to point out how awesome it is, given that we know that possessions matter, that Chris Paul is so good at a.) not turning the ball over and b.) ending the opposition’s possession via STL that he damn near averages more STLs than TOs, which is almost unfathomable. Stop trying to make it more than it is in your attempt to belittle any poster who dares present statistically bent posts that don’t comply with your world view.

      lol

      Jowles: “I posted his STL/TO numbers awhile back. Steve Nash’s STL/TO ratio is something like 0.25 for his career. Chris Paul? 0.95. Even Rondo, who’s a great defender, puts up a 0.72 for his career. No one’s even close to Paul in overall efficiency.”

    14. massive

      The steal percentage argument is pretty easy to understand. Not only is Chris Paul extremely efficient with the basketball, but he also creates possessions on the defensive side of the ball. He almost creates a new possession for every turnover he commits, which is incredible.

      Of course, I’m very biased (Chris Paul is tied with James Harden my favorite basketball players not on the Knicks), but I think using CP3s steal to turnover ratio to talk about his efficiency as a player is very fair.

    15. Juany8

      massive:
      The steal percentage argument is pretty easy to understand. Not only is Chris Paul extremely efficient with the basketball, but he also creates possessions on the defensive side of the ball. He almost creates a new possession for every turnover he commits, which is incredible.

      Of course, I’m very biased (Chris Paul is tied with James Harden my favorite basketball players not on the Knicks), but I think using CP3s steal to turnover ratio to talk about his efficiency as a player is very fair.

      I get that Chris Paul is efficient, but I don’t see how a steal to turnover ratio actually says anything about a player, especially when charges are also turnovers caused by an individual play, but they’re simply not kept track of at an individual level. It’s impressive that Chris Paul keeps his turnovers low, but that doesn’t have anything to do with a defense. It’s also nice to gain possessions by getting steals, but it doesn’t define a player’s defensive contributions pretty well, I’d rather have someone like Tony Allen who causes turnovers through intense pressure rather than someone like Allen Iverson just gambling a ton and mixing in steals with open paths to the rim. Chris Paul is still amazing regardless, I just don’t see the point in comparing his turnovers to his steals, especially since I’d rather have Rondo or Westbrook on defense over Paul, even if they’re getting less steals

    16. massive

      It’s not about defense, though. It’s about Chris Paul’s efficiency, and the stat should be used in addition to his other numbers. Chris Paul’s TS% is usually really high for a point guard, he rarely turns the ball over, and for every 10 Chris Paul turnovers, he has 9 steals. For every 100 turnovers, he has 95 steals. Basically, he makes up for an extremely high percentage of his turnovers by creating turnovers on the other side of the ball. I think that is it what THC was talking about when he mentioned net possessions.

    17. Juany8

      massive:
      It’s not about defense, though. It’s about Chris Paul’s efficiency, and the stat should be used in addition to his other numbers. Chris Paul’s TS% is usually really high for a point guard, he rarely turns the ball over, and for every 10 Chris Paul turnovers, he has 9 steals. For every 100 turnovers, he has 95 steals. Basically, he makes up for an extremely high percentage of his turnovers by creating turnovers on the other side of the ball. I think that is it what THC was talking about when he mentioned net possessions.

      I get “Net Possessions”, he’s just not even close to counting it right with the box score. Again, how is getting a charge not a net possessions gain (and Chris Paul is good at getting charges too!). If you play such fantastic defense that the opponent has to chuck up a prayer from 3, how is that not a net possession? If you cause a back-court/8 second violation, how is that not a net possession?

      As it is, the stat only compares turnovers to steals, which I find meaningless. It’s very possible to say that Chris Paul is both good at avoiding turnovers and getting steals without trying to turn their ratio of them into something meaningful. It’s an efficiency stat that makes Steve Nash look bad for crying out loud, and Steve Nash is one of the best offensive weapons of all time, both as a scorer and passer. He gets more turnovers by taking more risks than Chris Paul, but he makes up for it with extra assists, and of course he wouldn’t get too many steals. Seriously, any “efficiency stat” that makes Steve Nash look mediocre should probably be cast aside, it’s plenty possible to just analyze steals and turnovers as 2 entirely different categories.

    18. ruruland

      Juany8: I get “Net Possessions”, he’s just not even close to counting it right with the box score. Again, how is getting a charge not a net possessions gain (and Chris Paul is good at getting charges too!). If you play such fantastic defense that the opponent has to chuck up a prayer from 3, how is that not a net possession? If you cause a back-court/8 second violation, how is that not a net possession?

      As it is, the stat only compares turnovers to steals, which I find meaningless. It’s very possible to say that Chris Paul is both good at avoiding turnovers and getting steals without trying to turn their ratio of them into something meaningful. It’s an efficiency stat that makes Steve Nash look bad for crying out loud, and Steve Nash is one of the best offensive weapons of all time, both as a scorer and passer. He gets more turnovers by taking more risks than Chris Paul, but he makes up for it with extra assists, and of course he wouldn’t get too many steals. Seriously, any “efficiency stat” that makes Steve Nash look mediocre should probably be cast aside, it’s plenty possible to just analyze steals and turnovers as 2 entirely different categories.

      I sort felt like this didn’t need to be said but you did a good job with it anyway.

    19. The Honorable Cock Jowles

      Juany8: I get “Net Possessions”, he’s just not even close to counting it right with the box score. Again, how is getting a charge not a net possessions gain (and Chris Paul is good at getting charges too!). If you play such fantastic defense that the opponent has to chuck up a prayer from 3, how is that not a net possession? If you cause a back-court/8 second violation, how is that not a net possession?

      As it is, the stat only compares turnovers to steals, which I find meaningless. It’s very possible to say that Chris Paul is both good at avoiding turnovers and gett with extra assists, and of course he wouldn’t get too many steals. Seriously, any “efficiency stat” that makes Steve Nash look mediocre should probably be cast aside, it’s plenty possible to just analyze steals and turnovers as 2 entirely different categories.

      Because a FGA always has the potential of going in. It’s a possession used to make a scoring attempt, and is thus NOT a turnover. Over time, a bad team will have more of these “last ditch” scoring plays and have a consequently lower eFG%. A good defensive team will, over time, force more of these plays, and thus have a low team eFG% against.

      A turnover is a possession that is NOT USED IN A SCORING ATTEMPT. If there is no scoring attempt, there is a 0% chance of scoring points (outside of technical fouls and other events that happen so infrequently that they don’t make much of a difference in long-term results like eFG, TOV%, ORB% and FTM/FGA do). So we can separate TOV% and STL% completely from those “low percentage shots” because they ARE, in fact, accounted for in team defense.

    20. The Honorable Cock Jowles

      And also, the backcourt violation thing? That’s accounted for in opponents’ TOV%.

      Charges? Maybe statistically significant. I looked up some numbers, and this is my interpretation:

      In 2005-06, the Knicks were the worst team in net charges, having 118 net possessions lost to charges.

      http://www.82games.com/charge.htm

      If the Knicks were, instead, even on net charges, they would (reasonably) gain about 120 points according to their offensive rating. Since they scored about 7800 pts that season, charges would have accounted for 1.5% of their points scored that season. Do you find that statistically significant?

    21. ruruland

      The Honorable Cock Jowles: Because a FGA always has the potential of going in. It’s a possession used to make a scoring attempt, and is thus NOT a turnover. Over time, a bad team will have more of these “last ditch” scoring plays and have a consequently lower eFG%. A good defensive team will, over time, force more of these plays, and thus have a low team eFG% against.A turnover is a possession that is NOT USED IN A SCORING ATTEMPT. If there is no scoring attempt, there is a 0% chance of scoring points (outside of technical fouls and other events that happen so infrequently that they don’t make much of a difference in long-term results like eFG, TOV%, ORB% and FTM/FGA do). So we can separate TOV% and STL% completely from those “low percentage shots” because they ARE, in fact, accounted for in team defense.

      But TOV% and efg% allowed are not accounted for individual players.

      So, why are we simply focusing on STL % when there are far more plays an individual defender is involved in?

    22. ruruland

      The Honorable Cock Jowles: And also, the backcourt violation thing? That’s accounted for in opponents’ TOV%.Charges? Maybe statistically significant. I looked up some numbers, and this is my interpretation:In 2005-06, the Knicks were the worst team in net charges, having 118 net possessions lost to charges. http://www.82games.com/charge.htmIf the Knicks were, instead, even on net charges, they would (reasonably) gain about 120 points according to their offensive rating. Since they scored about 7800 pts that season, charges would have accounted for 1.5% of their points scored that season. Do you find that statistically significant?

      What’s the difference between charges and steals, statistically?

      What about turnovers that are neither steals or charges?

    23. The Honorable Cock Jowles

      ruruland: What’s the difference between charges and steals, statistically?

      What about turnovers that are neither steals or charges?

      The Knicks had 555 steals that season, and teams ranged from 822 to 508. So even the worst stealing team in the league had nearly four times as many steals as the worst team in the league in charges (-118).

      And turnovers that are neither steals nor charges? Accounted for in offensive TOV%.

    24. The Honorable Cock Jowles

      ruruland: But TOV% and efg% allowed are not accounted for individual players.

      So, why are we simply focusing on STL % when there are far more plays an individual defender is involved in?

      Because we have no way of differentiating between players in team defense without doing crazy backflips with +/- data that really doesn’t say much of anything. With steals, we see that the numbers 1) remain consistent over time, suggesting that individual players are mainly responsible for their own production and 2) the steals result in a turnover and thus end the opposition’s possession and begins one for the receiving team.

    25. The Honorable Cock Jowles

      ^^^ This is the best possible outcome for a team, short of a turnover that results in 2 shots and the ball.

    26. Juany8

      The Honorable Cock Jowles: Because a FGA always has the potential of going in. It’s a possession used to make a scoring attempt, and is thus NOT a turnover. Over time, a bad team will have more of these “last ditch” scoring plays and have a consequently lower eFG%. A good defensive team will, over time, force more of these plays, and thus have a low team eFG% against.

      A turnover is a possession that is NOT USED IN A SCORING ATTEMPT. If there is no scoring attempt, there is a 0% chance of scoring points (outside of technical fouls and other events that happen so infrequently that they don’t make much of a difference in long-term results like eFG, TOV%, ORB% and FTM/FGA do). So we can separate TOV% and STL% completely from those “low percentage shots” because they ARE, in fact, accounted for in team defense.

      What you’re ignoring here is the crux of the argument about some players getting steals by being aggressive instead of playing solid defense. You are assuming that a player only gets attempt to get a steal when he actually gets the steal. Like field goals, players miss on steal attempts, and it often results in a compromised defense. I’m sure this doesn’t apply as much to Chris Paul, but someone like Allen Iverson will rack up lots of steals cheating off his opponent and gambling, and there is some potential downside to each steal attempt, in that a miss could result in a very favorable look. A player like Ibaka has this exact same problem with Blocks, no statistic factors in how many times he allows an open layup by jumping every single time an opponent makes a shot motion.

      Basically, you’re giving credit for steals in blocks but not penalizing their misses. It would be like counting up field goals made and just ignoring the bad efforts, the team can get an offensive rebound anyways right?

    27. Juany8

      Juany8: What you’re ignoring here is the crux of the argument about some players getting steals by being aggressive instead of playing solid defense. You are assuming that a player only gets attempt to get a steal when he actually gets the steal.

      Wow that was some bad English, it should read, “What you’re ignoring here is the crux of the argument about some players getting steals by being aggressive instead of playing solid defense. You are assuming that a player attempts to get steals only when he actually gets the steal.”

      Site could really use an edit post feature lol

    28. njasdjdh

      Juany8: Wow that was some bad English, it should read, “What you’re ignoring here is the crux of the argument about some players getting steals by being aggressive instead of playing solid defense. You are assuming that a player attempts to get steals only when he actually gets the steal.”

      Site could really use an edit post feature lol

      Juany8, there’s not much in the way of evidence that Paul is that sort of player though. If you watch him you see that that isn’t the case. Even if you don’t, I feel confident that that’s the sort of thing Hollinger would have brought up in his Paul analysis (similar to how he killed Ibaka for the over aggressiveness problem) where instead he says:

      “Defensively, Paul was his usual pesky self. He led the league in steals per minute, and while his size is an issue, his ball hawking more than offset it. Opposing point guards had a 13.8 PER against him, according to 82games.com, and Synergy rated him well above average.”

      In addition, I’m not sure it was you, but the poster who thinks Russell Westbrook is a better defender than Chris Paul is insane. Unless, of course, what he meant was that Westbrook has better defensive tools/potential (which he does). Ironically, that same poster mentioned Rondo as another superior defender, and he is a guy that Hollinger cites as going for the steal too much, FWIW.

    29. The Honorable Cock Jowles

      The argument invariably becomes: “Because the data lacks some information, let’s disregard it.” I’m done playing that game.

    30. ruruland

      The Honorable Cock Jowles:
      The argument invariably becomes: “Because the data lacks some information, let’s disregard it.” I’m done playing that game.

      No one ever says disregard it. That’s you pouting at the end of every thread.

    Comments are closed.