Like us on Facebook


Follow us on Twitter





Results 1 to 13 of 13
  1. #1
    Join Date
    Aug 2008
    Posts
    102
    vCash
    1500

    Points per Shot vs. Points per Possession

    While reading the comments to the Lakers' game against Denver, I started wondering about a statistical question that has been bothering me for quite some time.

    It is generally assumed that the Points per Shot stat is a better way of assessing a players efficiency than his actual Field Goal Percentage. While that may be true, I find it flawed itself.

    This is not meant to bash Kobe, I just use him as an example. Kobe needed 32 shots yesterday to get 34 points. What I wonder about is this:

    Besides those 32 shots, Kobe actually forced the action on a few possessions more, but got fouled.

    Wouldn't be "Points per Possession", as it is already used for team performance, be even more accurate than Points per Shot?

    To underline my reasoning, I'll give you another example. Let's stay with Kobe.

    In the first season game against Houston, he shot 8 for 20 from the field and got 27 points. Thats 1.35 points per shot. But in this game, Kobe also took 12 free throws, making 11 of them. So in reality, he didn't just shoot 20 times, but probably more (considering that there might have been some and-1s or fouls after the team foul threshhold). More importantly, you can assume that he used 24-26 possessions. So even though he scored 27 on 20 shots, he actually scored 27 on 24-26 possessions, which is less productive.

    To put that into perspective, we have to take a player, who has a low FT%, like Dwight Howard.

    In his first game of the season, he shot 7 for 9 from the field and had 23 points, which would indicate a very productive performance with about 2.5 points per shot. However, in this game Howard also took 19 free throws, only making 9. Is that efficient? I don't think so. The box score says that Howard shot the ball 9 times when in reality he probably tried to shoot about 16-18 times. 23 points on 17 possessions would be a more accurate way of describing Howard's performance in this game IMO.

    For the team, it is actually completely irrelevant if a player goes down the court and shoots or if he goes to the basket and draws a foul. In both instances the team uses one of its possessions.

    What do you guys think?

  2. #2
    Join Date
    May 2010
    Location
    Central Europe
    Posts
    1,167
    vCash
    1500
    Yes Points per shot is pretty useless stat because it ignores possessions used by free-throws.

    Either you use True Shooting% if you wan't to know how effective someone is when he decides to Shoot. Or you take Offensive Rating for individual players if you wan't to know how much points he scores per possessions used (includes TO).

  3. #3
    Join Date
    Oct 2005
    Location
    Shakedown Street, Japan
    Posts
    30,250
    vCash
    1500
    lakers4sho posted the formula for possessions in his thread on the subject:

    POSS(x) = FGA(x) + 0.44(FTA(x)) - OREB(y) + TO(x)

    You can see here that each field goal attempt, turnover, and .44 of each free throw attempt, uses one possession, while each offensive rebound gains one back.

    In their last game against Denver (the one you're referring to I think, meiklwhite), Kobe had:

    32 FGAs
    13 FTAs
    1 ORB
    2 TOs
    34 PTS

    So if we look at straight points per shot (=34/32), he made 1.06 PPS.

    But if we look at points per possession (=34/(32+(0.44*13)-1+2)) he made 0.88 PPP, so we can see a pretty big difference there.

    BBR lists Kobe's career offensive rating at 112, which if I'm not mistaken means he's averages 1.12 points per possession (although their formula for calculating this may be more complicated). So if that's correct, we can see that even though he scored 34 points, he played well under his usual level of efficiency in that game.
    I blog basketball at Roundball Mining Company///Twitter: @denbutsu

    Atheists Of PSD

  4. #4
    Join Date
    May 2007
    Location
    California
    Posts
    27,992
    vCash
    500
    As Patman said, just use ts%. As of now, it's the ultimate scoring efficiency tool to use.

    Andrew Luck, Robert Griffin III, and Colin Kaepernick walk into a bar... To watch Russell Wilson win the Super Bowl.

  5. #5
    Join Date
    Oct 2005
    Location
    Shakedown Street, Japan
    Posts
    30,250
    vCash
    1500
    Quote Originally Posted by baller101200 View Post
    As Patman said, just use ts%. As of now, it's the ultimate scoring efficiency tool to use.
    Now that's a debate I'd like to see both sides of. I'm not qualified to say one is definitely "better" than the other as a measure of offensive potency (and I'd imagine the purists would say that neither is "better", but they both simply measure different things). But I'd still like to see the point/counterpoint of offeff vs. TS%.
    I blog basketball at Roundball Mining Company///Twitter: @denbutsu

    Atheists Of PSD

  6. #6
    Join Date
    Jul 2008
    Location
    Bushwood Country Club
    Posts
    58,297
    vCash
    1500
    I prefer to use TS% to evaluate a player's efficiency scoring the ball

    If you want the ultimate, you've got to be willing to pay the ultimate price. It's not tragic to die doing what you love.

  7. #7
    Join Date
    Oct 2005
    Location
    Shakedown Street, Japan
    Posts
    30,250
    vCash
    1500
    Quote Originally Posted by Hawkeye15 View Post
    I prefer to use TS% to evaluate a player's efficiency scoring the ball
    Why? And I'm not being facetious. I really would like to know.

    Case in point: Melo.

    He gets a lot of crap for forcing things too much and creating too many offensive fouls, which obviously count as turnovers.

    So why is it not better, when attempting to account for scoring "efficiency" (as an overall blanket term for being effective), to use a metric which does not take into account O-boards and TOs than one which does? He's (respectively) creating more or forfeiting opportunities for his team to score...
    I blog basketball at Roundball Mining Company///Twitter: @denbutsu

    Atheists Of PSD

  8. #8
    Join Date
    Jan 2005
    Location
    Miami, FL
    Posts
    28,039
    vCash
    1500
    Quote Originally Posted by DenButsu View Post
    lakers4sho posted the formula for possessions in his thread on the subject:

    POSS(x) = FGA(x) + 0.44(FTA(x)) - OREB(y) + TO(x)

    You can see here that each field goal attempt, turnover, and .44 of each free throw attempt, uses one possession, while each offensive rebound gains one back.

    In their last game against Denver (the one you're referring to I think, meiklwhite), Kobe had:

    32 FGAs
    13 FTAs
    1 ORB
    2 TOs
    34 PTS

    So if we look at straight points per shot (=34/32), he made 1.06 PPS.

    But if we look at points per possession (=34/(32+(0.44*13)-1+2)) he made 0.88 PPP, so we can see a pretty big difference there.

    BBR lists Kobe's career offensive rating at 112, which if I'm not mistaken means he's averages 1.12 points per possession (although their formula for calculating this may be more complicated). So if that's correct, we can see that even though he scored 34 points, he played well under his usual level of efficiency in that game.
    regardless of which stat you choose... the game is atrocious

    1.06 PPS = ugly game
    0.88 PPP = ugly game
    11/32 FGs (34%) = ugly game

    In most situations PPP is the better indicator...but in this case... it is a bad game however you slice it

    A better example would be for a stat-line like this:

    16 FGAs (10 FGs)
    20 FTAs (12 makes)
    9 TOs
    32 PTS

    In that case...

    PPS is going to say this is a GREAT game... 32/16= 2.00 PPS
    FG% is going to say this is a GREAT game... 10/16= 62.5%
    TS% even says this is a really good game... 64.5%

    BUT, PPP is going to tell a more complete story by looking including all the TOs....

    The PPP= 32 points / [16 + (.44)(20) + 9 ] = 32/33.8 = 0.95 points per possession (Ortg of 95)...which is anything but impressive...
    Last edited by daleja424; 11-15-2010 at 03:48 PM.

  9. #9
    Join Date
    Jan 2005
    Location
    Miami, FL
    Posts
    28,039
    vCash
    1500
    TS% gives you a good indication of shooting...but NOT of scoring efficiency...

  10. #10
    Join Date
    Jul 2008
    Location
    Bushwood Country Club
    Posts
    58,297
    vCash
    1500
    Quote Originally Posted by DenButsu View Post
    Why? And I'm not being facetious. I really would like to know.

    Case in point: Melo.

    He gets a lot of crap for forcing things too much and creating too many offensive fouls, which obviously count as turnovers.

    So why is it not better, when attempting to account for scoring "efficiency" (as an overall blanket term for being effective), to use a metric which does not take into account O-boards and TOs than one which does? He's (respectively) creating more or forfeiting opportunities for his team to score...
    very fair. Let me rephrase, TS% gives the best indication of scoring the ball, with all other indicators, and factors washed away. It simply tells us who scores the best. The fact that it doesnt involve TO's makes the overall scoring efficiency claim fishy at best.
    I think offensive boards can also have an effect to your scoring efficiency, if you create possessions/opportunities with that skill (Love).
    I think TS% is a great indicator of scoring efficiency, in the flow of an offense. Does it miss some things? Yep. But so does everything. PPS is not a proper tool compared

    If you want the ultimate, you've got to be willing to pay the ultimate price. It's not tragic to die doing what you love.

  11. #11
    Join Date
    Oct 2005
    Location
    Shakedown Street, Japan
    Posts
    30,250
    vCash
    1500
    Quote Originally Posted by daleja424 View Post
    TS% gives you a good indication of shooting...but NOT of scoring efficiency...
    Actually, I think I would frame it like this:

    eFG% = best measure of "pure" (, from the field, non-FT) shooting efficiency

    TS% = best measure of overall shooting efficiency (of all shots, including FTs)

    PPP -or- offensive rating (and correct me if I'm wrong, but these are essentially the same, right?) = best measure of overal offensive efficiency, including shooting, but expanded to include O-boards and turnovers (which add or lose offensive possessions) as well.
    I blog basketball at Roundball Mining Company///Twitter: @denbutsu

    Atheists Of PSD

  12. #12
    Join Date
    Jan 2005
    Location
    Miami, FL
    Posts
    28,039
    vCash
    1500
    I dont like it worded like that... b/c it doesnt take into account things like assists... and feeding someone the ball for an open look is extremely efficient.

    I think PPP or Ortg are not indicators of offensive efficiency, but rather of scoring efficiency... b/c it is only concerned with scoring rate...

  13. #13
    Join Date
    Jan 2005
    Location
    Miami, FL
    Posts
    28,039
    vCash
    1500
    For instance... which output is more efficienct offensively would you say?

    Player A:
    10-20 FGs (1-3 3FGs), 4 turnovers, 2 offensive rebounds, 10-12 FTs, 31 points, 3 assists

    OR

    Player B:
    10-20 FGs (1-3 3FGs), 4 turnovers, 2 offensive rebounds, 10-12 FTs, 31 points, 12 assists

    based on PPP, they were equally efficient... but i think the difference in 9 assists begs to differ


    PPP and Ortg apply specifically to how efficiently a person scores given acertain amount of opportunities with the ball in their hand...

    It is not a perfect stat... b/c we do not differentiate between TOs due to a poor pass and scoring attempt TOs... but what stat is perfect?!?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •