1. Football Picks Survivor is done for 2014. Congrats to billlaimbeer from The Pistons Forum

The Plus Minus: Addition by Subtraction

Discussion in 'Pistons and NBA' started by del, Jan 29, 2013.

  1. del

    del Team Captain Forum Donor

    Joined:
    Nov 13, 2009
    Messages:
    476
    Likes Received:
    269
    Gender:
    Male
    Location:
    Geneva
    I have seen it come up consistently in discussions on this board and used in places like the NBA Power Rankings (ie plus/minus studs), but seriously, does anyone really think this statistic is relevant?

    To me, this stat is more about who you are playing with, or against, than the player it is attributed to and can't imagine coaches or GMs care anything about it.

    I seem to think it must be somewhat relevant to have begun being tracked, so I may be way off and would be happy if someone can show me any articles on its use or how it is an effective measure.
     
  2. BallDon'tLie

    BallDon'tLie All-Star 3x Fantasy Champion

    Joined:
    Nov 5, 2005
    Messages:
    6,973
    Likes Received:
    3,226
    Location:
    Denver, CO
    coynejeremy and The Panda like this.
  3. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,836
    Likes Received:
    3,245
    Gender:
    Male
    Location:
    Chicaaago
    It is just regression analysis basically. So, the more different combinations that are included in the figure, the more meaningful it is. The more of a division between the starters and bench, the less relevant a comparison is between guys from each.

    The way I'm interpreting the Pistons' results from this year that shows Daye at the top and Villanueva not far behind is that our bench typically outplays the bench of our opponents while our starters typically get beat by the other teams' starters. So, pretty much every guy in our starting lineup will have his +/- figures crushed while every guy on our bench will have theirs in the positive range.

    Within our bench, I think that it is pretty clear that Drummond is the dominant force that makes us better than our opponents. However, Drummond mixes a lot of time with the starters and against other teams' starters. A guy like Daye pretty much only plays with Drummond and against other teams' bench players (except for his hot streak). Furthermore, a high +/- figure can be a sign that the player really fits in with a scheme.

    While the best players in the league typically lead their team in this stat, there are consistent exceptions (usually on teams that don't have a heavily winning record and have a stronger than usual bench).
     
  4. del

    del Team Captain Forum Donor

    Joined:
    Nov 13, 2009
    Messages:
    476
    Likes Received:
    269
    Gender:
    Male
    Location:
    Geneva
  5. del

    del Team Captain Forum Donor

    Joined:
    Nov 13, 2009
    Messages:
    476
    Likes Received:
    269
    Gender:
    Male
    Location:
    Geneva
    Thanks too TaShawn and that makes sense, but just can't help think there is too many other factors involved that make this a meaningless stat. Will skim through that thread BDL pointed out
     
  6. TaS

    TaS All-Star Forum Donor

    Joined:
    Sep 1, 2005
    Messages:
    17,836
    Likes Received:
    3,245
    Gender:
    Male
    Location:
    Chicaaago
    It's hard to ignore when you take the top 10 guys in the league (with your gut test) and then go and look at their +/- stats. Highly correlated.

    It's one complex data point. The purpose of it is ideally to extract the guys who make a positive difference in the game though a method that ignores all arbitrary box score stats.


    Pretty easy to call any singular stat meaningless and make a good case for that. However, you can bet that 30 out of 30 teams have at least one very smart guy behind the scenes taking a hard look at stats like +/- and beyond. They don't rely on what it says, but they take it into account and probably investigate further when it shows that something is out of whack. In my opinion, it is usually pretty easy to explain away strange +/- stats. For instance, if Lebron's +/- was lower than Bosh's, but Lebron played 48 minutes per night, I'd just suggest that the bench unit is weighing down LeBron's totals and that if you made Bosh play 48 minutes instead of him, it would reverse. Or if Kratksov led the Pistons, you could say, yeah, but that's only because he plays garbage time in wins and that only happens against very poor teams with poor benches, etc. But every once in a while, there are quirks that defy common perception. It is interesting when it leads you to rethink the total effect that a player has on the game. You might then key in on things like his passing that doesn't directly lead to assists, or his floor stretching ability, or maybe even his effective fouling.

    I think it is one of the essential stats to take into consideration, but I understand that it means different things depending on each team's situation. It means nothing to say that LeBron's +/- this year is higher than Michael Jordan's was in 1996.
     
  7. The Panda

    The Panda Garbage Post All Star Forum Donor

    Joined:
    Jan 29, 2012
    Messages:
    2,103
    Likes Received:
    1,214
    Gender:
    Male
    All stats in sports are highly correlated, it is essentially part of their nature, particularly in a sport like basketball. Think rebounds and winning. It doesn't make it meaningless.

    Just remember +/- is dependant on these other factors, so it good to used as relative information, rather than absolute. i.e if Drummond is -1 and Stuckey is -15, it doesn't make them both bad, it just means that a)the team probably lost and b) the bench unit was worse than usual. However, what it does tell you, is that relatively speaking, Drummond is a better on court presense than Stuckey.
     

Share This Page