The Stat Lab

Discussion in 'Pistons and NBA' started by TaS, Jan 6, 2009.

  1. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    I never know where to post statistical junk and I've been throwing it in the plus/minus thread, but that is kind of misleading for most of these things. So, I thought it would be a good idea to start a thread for miscellaneous statistical insights and random analysis.

    If you have something interesting or have a good idea for an area to research, throw it out here. Maybe we can figure some things out as a group and bring new angles to light.

    My intention is not to make this a boring math project or anything, so don't feel like you have to do rigorous analysis to post something. Sometimes, interesting stats are more about the creativity of what to look at rather than the number crunching.

    Here are some of the places that I use to dig in case you want to go mining.

    In depth coverage of the NBA from 82games.com (very in depth, takes some getting used to)
    PopcornMachine.net (great for game flow, +/- for each lineup combination and stint. Usually posts after midnight on game days).
    NBA.com: Stats Index (click on team, then player)
    Basketball-Reference.com - Basketball Statistics and History
    NBA Statistics and League Leaders - National Basketball Association - ESPN
    HoopsHype - The NBA Basketball Web Site (salary data)
    NBA.com: Statistics: AutoTrader.com +/- Stats (great for slicing up +/- combos from 1 player to 5 and under different timeframes).
    NBA Stats - CBSSports.com Pretty good filters
    NBA Dunks - CBSSports.com (dunks!)

    NBA Basketball Stats on TeamRankings.com
     
  2. roscoe36

    roscoe36 All-Star Administrator

    Joined:
    Jul 5, 2005
    Messages:
    20,566
    Likes Received:
    2,423
    Gender:
    Male
    Location:
    Dagobah
  3. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    I'll kick it off with a bit of a weird one.

    My hypothesis is that steals are usually more about chaos caused by overall defensive pressure than they are about picking pockets. When the offense gets rattled and when guys are covered or drive into dead ends, then bad passes are thrown and the defensive guards usually use their quickness to get the loose ball.

    So I looked at the player pairs data from 82games. For each player on the Pistons, I looked through the grid to see which one player was on the floor when they got their most steals. The pattern is fairly clear.

    Player/ "Steal buddy"
    Bynum/ Amir
    Iverson/ Amir
    Stuckey/ Brown
    Afflalo/ Amir
    Hamilton/ Afflalo
    Prince/ Amir
    Herrmann/ Amir
    Amir/ Sheed
    Dyess/ Afflalo
    Maxiel/ Amir
    Sheed/ Dyess

    So, of the 10 players not named Amir on the team, 6 of them happen to get their most steals when he is out there with them. Afflalo shows up twice and each big man, Sheed, Dyess, and Brown show up once.

    I think this is a partial explanation for Curry's belief that Amir does things that don't show up in his stats.
     
  4. Warthog

    Warthog Bench Warmer

    Joined:
    Sep 1, 2005
    Messages:
    2,587
    Likes Received:
    6
    Location:
    Sterling Heights, MI
    you actually missed an Amir, paired with sheed
     
  5. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    Yeah, but all that one means is that Amir gets more steals when he's paired with Sheed than he does with any other specific player. So that one is more a reflection on Sheed doing something well than Amir.

    It probably means that Amir is picking up some of those balls after Sheed pokes them loose.

    In other words, these are not the top 11 steal combination duos on the team (as the guards get credited with most of the steals), but rather I take each player individually on the left and point out which one player maximizes their steals. Whoever that player is, will be on the right side.


    A little confusing maybe, but I think it points something out. The Dyess effect (where it looks like he's stealing other players rebounds) made me think of this.
     
  6. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    BENCH SNAPSHOT

    Our bench. It may look weaker than it really is because we've had so many starters missing (and thus bench players promoted to the starting lineup... and thus players like Bynum, Herrmann, and Brown getting major bench minutes at times), but I want to point out a few things.

    San Antonio's bench players get 98 minutes per game and that leads the league.

    Detroit's bench players get 74.8. There are only 4 teams that play their bench less (ORL, LAC, ATL, NYK).


    However, the surprising thing is that our bench leads the entire NBA in terms of points scored per 100 possessions at 110.7 points.

    NBA Starter Reserve/Bench stats

    That is a very good indicator that we truly are deep and also have 2 distinct styles. While our starters are holding teams to very low FG%, or bench is coming in and making it a scoring fest.

    It will be interesting to see if these contradictory styles merge or diverge as the season goes on.
     
  7. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    STARTING LINEUPS VS FINISHING LINEUPS and MC's tendancy to go to small ball to finish games...
    1. We win the first quarter 70% of the time. That is tied with Cleveland for 1st in the NBA.
    2. We win the 2nd quarter 42% of the time. Bottom 25%.
    3. We win the 3rd quarter 54% of the time. An echo of the first quarter probably, because it's the same players, but they come out sooner, so the effect is smaller.
    4. Here's the bad part. We only win the 4th quarter 36% of the time. We're basically tied for last with the Clippers and the Wizards as they are at 35% and 34%.
    Since we have the talent to start each game as strong as any team in the NBA, the question is, why are we sucking so badly at the end of games? You may be fooled by games where we pull it out in the end, but forget that we started the 4th quater with a significant lead. We haven't been blowing teams out like we used to do, and this is the reason.

    What can MC do to correct this problem? I'm guessing that it may be small ball related, but can't prove it.
     
  8. coynejeremy

    coynejeremy All-Star Administrator 1x Fantasy Champion Forum Donor

    Joined:
    Oct 9, 2005
    Messages:
    6,255
    Likes Received:
    1,158
    Location:
    Georgetown, KY
    As far as I remember, we have been getting outscored in the fourth quarter for years. It's something that I almost always seem to notice right before reading the recaps on espn.
     
  9. Delfino Delivers

    Delfino Delivers Bench Warmer

    Joined:
    Feb 26, 2006
    Messages:
    1,406
    Likes Received:
    0
    Location:
    Michigan's Thumb
    Could it be too many minutes in the first and third quarters for the starters while the opposition is getting more rest for the kick at the end of the game? Part of the reason I think we win first quarters is because our starters usually play the whole first quarter and some of that time is against the oppositions bench.

    Our last two coaches have IMO, played the starters too many minutes to start the games. I would like to see subs a little sooner in the first quarter and the starters minutes spread out more.
     
  10. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    We used to have higher win margins and I think that GT may be part of the explanation for poor 4th quarters over the past few years, although I didn't know they were poor.

    It seems like we have had 1 or 2 games this year where GT was in play. Think about how rare Sharpe played for instance. In prior years, there were those Darko GT showcases where he had to remove his earrings, put out his cigarette, get off the phone, and enter the game in the 4th quarter.
     
  11. BallDon'tLie

    BallDon'tLie All-Star 3x Fantasy Champion

    Joined:
    Nov 5, 2005
    Messages:
    6,807
    Likes Received:
    2,925
    Location:
    Denver, CO
    That's good stuff TaS!
     
  12. pass99

    pass99 All-Star Forum Donor

    Joined:
    Aug 6, 2006
    Messages:
    2,728
    Likes Received:
    402
    Oh,Oh...and I thought we have seen and lived through the last eight years of this form of cognitive analysis. :bolt:
     
  13. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    Who's the real BIG MAN?
    (all on a per 40 minutes basis).

    Dunks:
    1) MAXIELL- 2.3
    2) Amir- 1.5
    3) Brown- 0.9
    4) Prince- 0.6
    5) Sheed- 0.4
    6) Dyess- below the radar

    Defensive rebounds:
    1) DYESS*- 10.8
    2) Sheed- 7.9
    3) Brown- 7.5
    4) Amir- 6.5
    5) Prince- 5.7
    6) Maxiell- 4.5
    * keep in mind that he steals rebounds

    Offensive rebounds:
    1) AMIR- 5.2
    2) Maxiell- 4.4
    3) Dyess- 3.5
    4) Brown- 3.2
    5) Prince- 1.9
    6) Sheed- 1.0

    Blocked shots:
    1) AMIR- 2.7
    2) Maxiell- 2.3
    3) Sheed- 1.9
    4) Brown- 1.4
    5) Dyess- 1.1
    6) Prince- 0.8

    Even though Prince is not a big man, I threw him in here because of his play at the PF spot on occasion. He's a good water mark for these guys. If you're a big man and you're below Tay in one of these stats, then your sort of soft and need to man up in that area.

    Brown is very respectable in all of these categories. Overall, I'd say that he was a solid pick-up for the money and is filling a valuable need for us against bigger opponents.

    Sheed is still on pace to extend his 100 3's + 100 blocks streak. It's pretty obvious that Sheed's offensive rebounding is way below par, but that's just because he's not in the neighborhood.

    Maxiell and Amir show that they are the same player in many ways with the blocks, dunks, and offensive rebounding. But as you can see, Maxiell embarassingly falls short of the Tay water mark on the defensive glass.

    And Dyess is adding more that I would have expected on the boards, especially defensively. It makes sense that the vets use their position well to secure these while the youngsters use their energy and athleticism to steal offensive rebounds.


    Overall, this is a very good and diversified stable of bigs. When you also consider Sheed's 3-point range and overall defense, Dyess's mid range jumper, Amir/Maxiell's defensive speed, and KB's bulk and relatively quick feet, then you have a unique group here. No Dwight Howard, but still a viable package.
     
  14. TheeTFD

    TheeTFD All-Star

    Joined:
    Sep 1, 2005
    Messages:
    9,810
    Likes Received:
    571
    Gender:
    Male
    Location:
    Los Lunas, NM
    Sheed gets most of his steals when paired with Dyess.
    Interesting.
     
  15. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    Surprising problem: Offense
    • This season to date, Det has the 6th best PPG allowed at 93.5. We have the 7th best opponent FG% at .443. Of the top 6 defensive teams, our point differential is easily the worst, at only +0.97.
    • Over the past 10 games, whether it's because of weak opposition or improvement, our defensive stats have risen to #2 in PPG allowed and #1 in opponent's FG%. If you look at the last 5 only, then it gets even better and we are #1 in both categories. :winner_first_h4h:
    So, what's the problem?

    It's our offense, which checks in at 26th in the league, pinched between the OKC Thunderclaps and the Memphis Darkos. You might argue that this is because of our slow pace, but that same pace was is built into those defensive stats above. Besides that, our pace is actually 2 possessions faster per game that it was last year, from 87 to 89.

    What if you argued that it was our early season struggles with the loss of CB and the integration of AI. We've been scoring more lately since Stuckey came alive, right? Well, over the past 10 games, our offense has actually been the 2nd worst in the NBA, ahead of only Washington. :nopompom:

    I'm here to point out that it isn't so much our offense as you would imagine it, but more of a loss of possessions, which means that our offense doesn't get as many chances.

    So, here's what it boils down to as far as I can tell:
    1. Rebounding. Surprisingly, our rebounding has actually gotten a little worse this year. The main difference is that our offensive rebounding % has dropped from 31% to 27%. It is still a good percentage league-wide, but the loss in extra possessions from last year combined with our lower FG% when we do get it works out to a loss of 3.2 points.
    2. The free throw gap has widened. Last year, we lost an average of 1 point per game because our opponents made 1 more FT than we did. This year, we are making 1 less than last year and our opponents are making 1 more. So, we've taken a previous slight weakness and made it worse on both ends of the floor. In terms of us fouling the other guys, that doesn't matter that much, because our defense is still holding the opposition to a level which makes us an elite defensive team. But what is the explanation for us getting fewer FT's on offense? Stuckey + AI doesn't result in more FT's? This is a mystery to me. Anyways, this problem is easy to quantify. Loss of 2 points per game, net.
    3. Turnovers- These last 2 are very minor effects, but I'll point them out to explain the remaining difference. Our turnovers are slightly higher this year on offense, yet we are causing the same amount on defense. Call it 1 turnover per game difference, which = about a loss of 1 point. Not a big deal given the loss of a PG who was considered a master of protecting the rock.
    4. Shooting- This one is a little complicated. Our overall FG% is nearly identical to last year at .456 vs .458. When you adjust for 3-pointers, then our eFG is marginally worse at .488 vs .495. Still not a big deal. Even so, it works out to a loss of 1 point. But the interesting thing is how we are getting our points. We are shooting almost 2% worse from the perimeter, but > 6% better from the interior. Our jumper frequency has gone down from 71% to 68%, which I think is a welcome change. However, like all teams, we still take the majority of our shots from outside the paint, and the lower % there outweighs our improvement on the interior.
    And that right there explains the errosion of our point differential.

    The causes of these changes are not simple and the solutions are even more difficult to find. The rebounding issue is particularly perplexing to me. It seems like we've been grabbing a ton of offensive boards. Is it just that we're noticing it when new guys do it and we're not seeing the decline in the one's towed down by the rest of the team... say by the guards?

    The foul problem is equal between drawing and giving, so this is a tough fix. Our guards are already penetrating more, so I don't think they can do much about this. Having Sheed and Dyess draw more fouls from the post is the only realistic solution I think, but even that is a pipe dream.

    In terms of giving fouls, Amir is the obvious culprit, and AI's D is probably the intangible culprit for a portion of it (by not containing the perimeter). But if you limit his play, you give up in other areas. His +/- stats are showing that he's a net benefit, so if you fix the fouling problem by taking him out, you create new problems for the team that would pop up in new categories on my board above.

    Is the small ball effect embedded in these numbers? That might explain the rebounding mystery.

    It's tough to make any real generalizations here because the team has had no rhythm this year. CB and Dyess left. Will Bynum played a bunch. AI got thrown in the mix. Dyess came back. There have been significant game losses due to injury. MC is dabbling quite a bit and may even been blindly experimenting.

    So, I'm not going to conclude anything here, but just hope to show some things that are and aren't working.
     
  16. dba

    dba All-Star Moderator 1x Fantasy Champion Forum Donor

    Joined:
    Jan 24, 2006
    Messages:
    4,213
    Likes Received:
    362
    Location:
    Ann Arbor/Chicago/Walland, TN
    So, if AJ fouled less, say two fewer a game, that would make up the FT differential, and since he wasn't in foul trouble, would play twice as many minutes, easily making up the offensive board discrepancy, and probably scoring more points than Max or Kwame.

    Clearly, it's all Amir's fault.
     
  17. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    His fouls aren't all resulting in FT's and the guys he's fouling tend to be the worst FT% guys, so it would probably have to be more like 3 fouls less. Since he only averages 3.5 fouls per game, that is asking him to cut his foul rate by 86% and become as clean on D as Wilt Chamberlain.

    But, of all the problems I've identified, I think the easiest to fix will be the elimination of SB (check), and more playing time for Amir through the use of mental discipline and physical restraint on his part.

    All of the other problems are intrinsic and we'll have to improve in other areas.
     
  18. roscoe36

    roscoe36 All-Star Administrator

    Joined:
    Jul 5, 2005
    Messages:
    20,566
    Likes Received:
    2,423
    Gender:
    Male
    Location:
    Dagobah
    Hypnosis.
     
  19. TheeTFD

    TheeTFD All-Star

    Joined:
    Sep 1, 2005
    Messages:
    9,810
    Likes Received:
    571
    Gender:
    Male
    Location:
    Los Lunas, NM
    Well Curry has known all along Amir helps us win. I think now he is dedicated to those ends.
     
  20. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,638
    Likes Received:
    2,972
    Gender:
    Male
    Location:
    Chicaaago
    Here is one more deep thought on the Amir fouling issue.

    Besides Amir getting in foul trouble and not being available for more than 25 MPG, how does his fouling hurt us?

    So, what I want to look at is the difference between guarding a guy normally and fouling him.

    I'm going to start out with an assumption that isn't 100% true, but I hope that it is 90% true. The assumption is that when Amir commits a shooting foul, it is usually in the paint and usually against the other team's PF or their C.

    Let's look at typcial big men's shooting % in the paint vs. their free throw percentages. I'll just use our 5 big men as a representation for the league as a whole to approximate.

    Player/ inside FG%/ FT%
    Dyess/ .655/ .667
    Amir/ .649/ .750
    Max/ .693/ .536
    Sheed/ .579/ .762
    Brown/ .633/ .413

    AVE/ .644/ .626

    Now I realize that this is just an approximation, but in general, big men shoot in the mid 60's both in the paint and at the line as a whole.


    So here are the scenarios to consider:

    1) Amir guards the guy as hard as he can and doesn't foul. In this situation, Amir will greatly reduce that player's FG%. Say he will reduce it from .644 to .444. There is actual data out there that shows the different shooting percentages between wide-open, open, contested, and heavily contested. I looked at how it affected big men only and used the average drop between categories.

    In this situation, the expected value of the player's FGA will be roughly 0.88 points.

    2) Amir guards the guy as the average NBA big man would, trying to affect the shot, but careful not to foul. For this scenario, I'll assume that the opposing player averages the same as our big men, and shoots .644.

    In this situation, the expected value would be 1.29 points.

    3) Amir gets carried away and fouls the other player in the paint on a post move, an alley oop, a slash, dunk attempt, or any other typical paint FGA. In this situation, the player will get 2 free throws. In this case, the expected value would be 1.23 points. We need to factor And1's into this to make it complete though. At the peak of his career, Shaq was getting and1 opportunities on 8.5% of his FGA's. Let's be on the safe side and assume that Amir fouls on a made basket at that rate. So, we're going to tack on .085*2.626 on to the expected value. That equals .31, so we'll increase the expected value to 1.54 in this situation. But now we have to adjust the 2 free throws down, b/c we're only saying it happens 91.5% of the time with and1's happening the other 8.5%. This reduces the number by 0.11 and the net result is an expected value of 1.43 points.


    In conclusion, situations where Amir plays smart conservative defense on his man will work out slightly better for the team as when he plays crazy D and fouls the bejezus out of his man in the paint. Where he really helps us is in the first scenario where he plays crazy D and gets away with it. The more he can change scenario 3 without also changing scenario 1, the better off we'll be.

    But after going through this, I feel fairly confident in saying that his style of play and his penchant for fouling doesn't hurt the team, his absence on the court hurts the team. In other words, if they eliminated the 6 foul limit, we should have no problem with his style as it would be a good gamble.

    Cutting out the obvious and1's, like the one against Oden would really help though.


    * just want to acknowledge the point that was brought up earlier about his non-shooting fouls hurting the team at the end of quarters because of the bonus. It is a small factor, but I don't think it is calculatable by me. Also, the rule about the 2nd foul under a minute putting a team in the bonus negates a big chunk of this potential effect.
     

Share This Page