Can we define consistency?

Discussion in 'Pistons and NBA' started by TaS, Oct 23, 2007.

  1. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    OK, this was not easy.

    I figured out how the Pistons' players perform when they are given different minutes.

    I'll spare you the calculation and just show the results:

    As a whole, players on the team perform 12% worse than the overall average when they are given 10-20 minutes per game (remember, I previously excluded <10 minute data).

    They perform 5.9% better when given 20-30 minutes

    5.88% better when given 30-40 minutes

    and 1% better than their average when given 40+.


    So, when you graph it, you see an arc. Basically, players don't play well unless they get 20+ minutes. Once they play more than 40, they start to decline.

    Not a single player on the team did better than their average with less than 20 minutes. Not even Dyess or Webber (older players).

    I'll go through the players and point out their best time slot:

    Nazr 20-30
    Dyess 30-40
    Chauncey 40+ (gets stonger and stronger)
    Sheed 30-40 (significant drop off after 40, gets worn out playing d?)
    Webber 30-40 (drop off after 40 minute mark)
    Rip 20-30 (energy player?)
    Maxiell 30-40
    Delfino 30-40
    Davis 20-30
    Prince 30-40 (dropoff after 40)
    Hunter 20-30
    Murray 30-40 (he shows biggest improvement on team in this range)
    Blalock (he only played < 20)

    So, for the most part, the more you play a guy, the better he does per minute, until you push him to the brink of exhaustion (except for Chauncey). Maxiell and Murray both responded very positively when given more minutes. Also, Webber and Nazr had a lot of trouble with < 20 per game.

    As far as who we should worry about Flip "overplaying", I guess I would say 1) Rip, 2) Sheed and T2) Prince. Don't worry about Chauncey.
  2. linwood

    linwood All-Star

    Joined:
    Sep 2, 2005
    Messages:
    4,382
    Likes Received:
    1,383
    Gender:
    Male
    Location:
    Seattle, WA
    Can I check out the calculation?
  3. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    I took each players individual game data and added a column for each games per 40 minute production stat. Then I sorted all of the rows by minutes played. Then I found the average "production" for the games in each bucket (10-20 minutes, 20-30 minutes, 30-40 minutes, and 40+ minutes).

    Then I figured out how much below or above the players total production average (as a percentage) occurred for each bucket.

    I also did a calculation where I summed up all of the bucket variances to find the overall trend for players in general. That is what produced the arc that showed that players perform better when they get 20-40 minutes than less or more.


    However, keep in mind that this is a correlation, and not necessarily a cause-effect. It is entirely possible that the reason players get < 20 minutes is because they are playing poorly. So the cause effect might be the opposite of what I'm contending.

    Also, it may be a little bit of both. Players get taken out when they play poorly, and they tend to play better when they are allowed to get into a flow.
  4. dba

    dba All-Star Moderator 1x Fantasy Champion Forum Donor

    Joined:
    Jan 24, 2006
    Messages:
    4,201
    Likes Received:
    349
    Location:
    Ann Arbor/Chicago/Walland, TN
    True across years, but that's not the point. The idea is not to try and project how a player will do next year or in three years - how they will develop. The idea is to find a fair way to equalize minutes so that stats can be compared for players playing differing numbers of minutes right now.
  5. buddahfan

    buddahfan Retired from Forum

    Joined:
    Sep 1, 2005
    Messages:
    5,361
    Likes Received:
    1
    Its too complicated and can't be done without a computer program that analyzes all the variables on the floor. Using box scores won't give you correct numbers.

    Players who play fewer minutes usually play against inferior talent, so you have to adjust their numbers for who is guarding them.
    Numbers must be adjusted for defensive schemes used, because a team might zone your team because of other players on the court on your team. This can lead to different set of numbers than if your opponents were playing man to man. In fact you need to adjust offensive numbers for all types of defensive since different defensive schemes favor some players and hurt others.
    You also have to take into account the game situation
    Also the players on ones team on the floor with an individual must be taken into account.
    Home and away
    Officials
    Opponent - Some teams play better defense some worse.

    In addition to the above there are many other variables that need to be considered.

    Its like chasing windmills.

    :hoops:
  6. dba

    dba All-Star Moderator 1x Fantasy Champion Forum Donor

    Joined:
    Jan 24, 2006
    Messages:
    4,201
    Likes Received:
    349
    Location:
    Ann Arbor/Chicago/Walland, TN
    Isn't that why they build them?
  7. BillLaimbeer

    BillLaimbeer All-Star 4x Fantasy Champion Forum Donor

    Joined:
    Sep 7, 2005
    Messages:
    14,974
    Likes Received:
    3,035
    Location:
    Michigan
    If you build the windmills, they will come.
  8. buddahfan

    buddahfan Retired from Forum

    Joined:
    Sep 1, 2005
    Messages:
    5,361
    Likes Received:
    1
    Wind power is experiencing substantial growth in various parts of the word.

    [​IMG]

    Wind power - Wikipedia, the free encyclopedia


    Of course as Pistons fans we really love this kind of windmill best. LOL

    YouTube - Amir Johnson DUNK!


    :hoops:
  9. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    What I did should have taken a lot of that into account. I looked at how Rip played for example, when he got 20 minutes, 25 minutes, 30 minutes, etc. I doubt that the quality of competition varied for these different games. He was probably matched up against the other teams starters for the majority of the time. It was a pretty large data set. With him, I think it was 90 games or so, including the playoffs.

    Rip shows decline after 30 minutes per game. I'm not saying that they shouldn't play him more than that, because a declining Rip is still better than our next best option. But if you have control of the game, he would be a good guy to rest for the next one if he has already played 30 minutes.

    And it is pretty obvious to me by looking at the data that players just don't play their best if they get less than 10. If there is any sport where you need to be warmed up, it is bball.
  10. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    I may as well just post all of my obscure thoughts under this thread.

    I decided to rank the Pistons in order of density (Weight per vertical inch).

    Using the "official" measurements from the Pistons' website, here is how it looks (except, they had 245 for Samb, which is obviously a mistake... so I took 215 off wikipedia for him):

    1) Jason Maxiell- 3.29 lbs/inch
    2) Mohammed- 3.05
    3) McDyess- 3.02
    4) Hayes- 2.81
    5) Afflalo- 2.79
    6) Wallace- 2.77
    7) Billups- 2.69
    8) Stuckey- 2.66
    9) Prince- 2.65
    T9) Dupree- 2.65
    10) Hunter- 2.64
    11) Murray- 2.63
    12) Johnson- 2.59
    13) Mejia- 2.56
    14) Samb- 2.53
    15) Hamilton- 2.44


    So by my calculations, Samb would need to put on 25 pounds to be as dense as Hayes.

    Maybe a good fast break lineup would be all of our feather weights?
    Murray
    Rip
    Mejia
    Johnson
    Samb
  11. coynejeremy

    coynejeremy All-Star Administrator 1x Fantasy Champion Forum Donor

    Joined:
    Oct 9, 2005
    Messages:
    6,144
    Likes Received:
    1,026
    Location:
    Georgetown, KY
    So does the fact that Chauncey is a 40+ player reflect his clutch ability? Is his ability to make three-pointers when the game is on the line (ie: when his minutes played has gotten to or past the 40-minute mark) causing his consistency to improve to the 40+ level?

    I really like the density calculations. I don't think anyone has ever done that before. :)
  12. buddahfan

    buddahfan Retired from Forum

    Joined:
    Sep 1, 2005
    Messages:
    5,361
    Likes Received:
    1
    Body fat is very relevant when looking at a person's weight relative to their height

    Weight management and the athlete

    bodyfattest.com - What Should Your Body Fat % Be?

    continued on link

    Bodybuilding.com - Alan Stein - Power: How Basketball Players And Other Power Athletes Can Learn To Jump Higher And Run Faster!






    :hoops:
  13. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    Chauncey's production came primarily from points and secondarily from assists.

    Here is a breakout for you (per 40 minutes).

    40+ minutes of PT
    18.92 points
    7.72 assists
    3.84 rebounds

    30-40 minutes
    18.52 points
    7.28 assists
    3.48 rebounds

    < 30 minutes
    15.96 points
    6.36 assists
    3.34 rebounds


    My theory is that he is the only player to show improvement over 40 minutes because:

    1) in a close game, Flip keeps him in, and then Chauncey is the one who shoots all of the free throws if we happen to have a slight lead and the other team resorts to fouling.

    2) Same reason as above, but Chauncey draws fouls himself.

    3) He heats up with his jumper the longer he is out there

    4) Since he is our go-to player at the end, he is also able to get some easy assists when they key on him.
  14. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    Just want to give a final summary for our preseason. I'm excluding everyone who played less than 50 total minutes (Lindsey, Fitch, Johnson). These are all per 40 minutes. I could have done per minute, but these are more understandable.

    Scoring-
    - Chauncey, Hayes, and Stuckey were the studs with 21-19 points.
    - Then, there was a whole pack of guys in the 16-18 point range (JMax, Prince, Murray, Sheed, Afflalo, Rip).
    - The bottom 5 were clearly a notch back in the 7-10 point range (Mohammed, Samb, Dyess, Dupe, and Mejia).

    Rebounding-
    - Maxiell was the king with 12.6
    - 2nd tier was in the 8.5-10 range (Sheed, Dupree, Mejia, Nazr, Dyess, Samb)
    - Everyone else was further behind in a similar range.

    Assists-
    - Chauncey was far and away the best at 9.6
    - Murray was a clear second with 7
    - Rip, Stuckey, Mejia, and Dyess were next in line with 5-4 a piece.
    - Maxiell and Samb were last... as you would expect.

    Steals-
    The "all hands" team-
    By some weird freak of nature, Nazr was the leader!
    Stuckey was 2nd. Billups, Mejia, and Murray were the next closest.

    Blocks-
    Again, Nazr dominated with 4.5 blocks.
    Samb and Sheed were next with about 2.5 a piece, followed by Prince and Maxiell with < 1.5.

    Assist to TO ratio-
    I'm only ranking the guards here, b/c this is really a guard stat.
    Rip 5:1
    Chauncey 3.6:1
    Murray 2.4:1
    Stuckey 1.1:1
    Afflalo 0.47:1

    Clearly, there is a learning curve in this department.

    Personal Fouls-
    Maybe this is why Flip was getting upset with the unfair rookie calls...
    most foul prone = Samb with 6.7
    Next in order was Dupree (6.2), Mejia (5.2), Maxiell (5.1), Stuckey (4.7), Nazr (4.5), and Afflalo (4.4).

    The vets were all in the 1-4 range.
  15. dba

    dba All-Star Moderator 1x Fantasy Champion Forum Donor

    Joined:
    Jan 24, 2006
    Messages:
    4,201
    Likes Received:
    349
    Location:
    Ann Arbor/Chicago/Walland, TN
    Lots of mentions of Mejia everywhere except scoring.
  16. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    I didn't pay attention to him in college and can't really tell what his niche can be. Hopefully he comes into his own in the D-League.
  17. buddahfan

    buddahfan Retired from Forum

    Joined:
    Sep 1, 2005
    Messages:
    5,361
    Likes Received:
    1
    Mejia and Dupree both shot around 25% for the pre-season games.

    :hoops:
  18. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    Of all the players I included here, Rip, Mejia, Dupree, and Nazr had the fewest minutes. So Mejia was 4-17 shooting while Rip was 7-18. These are not big enough sample sizes to go by. Kobe could easily shoot that many shots in one half.

    I would guess that the small sample size is the reason for Nazr's rebounding, and for sure his steals, being so high. But still, he didn't get many minutes and he over achieved during those. Also, the small sample probably makes Mejia look better than he really is in some of those categories.

    Everyone else had enough minutes to get a general idea of what their strengths are, I think.


    We'll see how their real stats compare to this sampling later in the year.
  19. professor

    professor Bench Warmer Forum Donor

    Joined:
    Nov 17, 2005
    Messages:
    1,030
    Likes Received:
    2
    Location:
    Oberlin
    first off, i'm just amazed by how you guys are thinking: inventing ways to use statistics to shape/correct perception. wow.

    second, with respect to the list below. interesting. but it made me wonder how these things play out over the long haul, esp. chauncey. that is, chauncey gets stronger the more minutes he plays in a game. but does he maintain this quality after a full season of playing those minutes?
    just a question
  20. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,402
    Likes Received:
    2,760
    Gender:
    Male
    Location:
    Chicaaago
    I think that it is too difficult to find that out, because it is hard to isolate the variables of him gaining experience and also aging. Those are huge variables and the marginal difference between getting 35 mpg in 2006 vs. 41 mpg is probably going to be lost in the greater noise.

    However, a shorter term way to look at it would be to analyze the splits for how the player performs with 0 days rest, 1 day, 2 days, etc. Since they keep track of that, it is easy to do.

    Putting my same formula in place, Chauncey did not show a real pattern in 2006. He played best with 3+ days rest, but that data set only had 69 minutes (insignificant). After that, his best category was 0 days rest. Then it was 2 days rest. Then 1 day. Then 3 days.

    I guess my overall conclusion is that since he basically kicks butt in the +40 minutes category and the 0 days rest category, that he is just a freak.

    However, Rip seemed to be affected most adversely by excess playing time in a game, but he also performed best with 0 day's rest, followed by 1 day's rest. So, maybe he needs to play often to have his shot be "on", but his legs get tired with all the running around in each game??

    If this were true, then players who depend on making shots might fit this pattern, while players who do hard work (like Ben Wallace) might show the opposite.

    -- I just checked Ben's stats, and he also performs best with 0 days rest.

    Who knows. Maybe the bigger pattern is the same as the in game pattern. Players play better when they play frequently, but there will be a drop off at a certain point if it is pushed to the extreme.

Share This Page