The merits of the often used statistic of player +/-

[indeedproceed]

You know, PPFGA is a stat that you'd think you'd see more often, but actually never do.

[mmmmm]

You know, PPFGA is a stat that you'd think you'd see more often, but actually never do.

Probably because it hides the defensive benefit of turnovers and other events that terminate possessions that don't generate an FGA.  You could generate 3 pts on every FGA, but if you turned the ball over 50 times (obviously exagerating)that would tend to be hidden by PPFGA.

Points per possession incorporates all those effects.

On the topic of this thread: +/- and it's derivative stats are best used to understand the effectiveness of 5-man units over time.  Individual ratings have to be looked at with consideration to context.  In a given game, small differences are too easily within the noise of a basket or two.  However, Ray's +22 for  44:58min (and the -18 for the 5:02 he was 'off'!!!) is so far above the game margin that it suggests that his mere presence on the court probably did have a real positive effect.   

You can't just be standing around for 45 minutes doing nothing on an NBA court because that means your guys are playing 4-on-5.  You cannot usually out-score the other team by +22 points in those minutes playing 4-on-5.

+/- doesn't tell you what _exactly_ was happening during the minutes the player was on the floor.  But they can, over time, roll up the net value of everything they are doing, whether it is making shots or grabbing rebounds or doing non-stat things like setting good screens, making smart (non-assist) passes or spreading the defense or playing good defense on the other end.   Guys like Perkins can post almost no counting stats and still post a good +/- for a game if they play good defense and set some good hard picks while on the floor.

I wouldn't read too much into it for any single game other than "He probably played a good game away from the ball."

In Ray Allen's case, we have a lot of history that shows that team's often game-plan to try to take him away from the offense and that he does have an effect on defensive spacing with his threat to shoot and his movement without the ball.  So if one went back and analyzed the game in detail, you might look for cases where he pulled more than one defender to a side, or someone else's man off a screen, trying to deny passes to him while instead Pierce or KG ended up free to get the ball or Rondo feinting to Ray and then slashing instead. 

There are all kinds of things that players do that are 'good' or 'bad' that don't show up in a stat sheet.   

Again, +/- doesn't tell you what those things were, but it does add up all their net effects on the scoreboard - which is the most important stat of all.