Does That +/- Stat Mean Much To You?

[ABM]

When I peruse the Blazer boxscores, my eye always seems to catch that particular stat. I've been noticing these days that Jerryd's +/- numbers appear relatively positive.

Is there much to draw from those, though? I realize that stats, alone, can sometimes be misleading.

 

[Nikolokolus]

raw +/- probably isn't as valuable as adjusted +/-, but it can give you a glimpse of the kind of defensive impact a player has on the court when looked at over a larger sample size than a single game. But in tonight's case I think you could say there was very strong correlation between his defense and his +/- stat -- Stuckey was turning the ball over and missing shots when Jerryd was on him, and at the other end Jerryd ran the break with LMA, got a nice dunk off of a steal and got to the foul line 4 times in very limited minutes.

 

[Mediocre Man]

Overrated stat

 

[The Sebastian Express]

It's a statistic that shouldn't be taken at face value. It doesn't tell you when just looking at it if that person is a minus because the four other guys on the court absolutely sucked, what unit on the other team they were playing against, how many times that player touched the ball offensively, how often that player's teammates failed to grab rebounds or how often they got burned on defense. It doesn't say how much of an affect that guy had during a team's run, did they contribute to anything positive or was it the four other guys on the court? Much like all stats this is a misleading one and you're often left best deciding by watching the games with your own eyes.

For instance I do not need a plus/minus to tell me Jerryd played good ball last night or some of the nights before.

 

[Wizard Mentor]

+/- during a single game may indeed be an overrated stat, but +/- over the course of 30 or so games is meaningful.

 

[BLAZER PROPHET]

To me it's meaningless.

As are most stats.

 

[Wizard Mentor]

To me it's meaningless.

As are most stats.

You could even argue #wins is meaningless (based on luck and officiating), but what then has meaning?

 

[hasoos]

To me it's meaningless.

As are most stats.

Wow what a generic blanket statement. Lots of stats are useful. +/- for me is a good indicator of how the unit functions while a player is on the floor. It is not the whole picture. But it gives you an idea of how it was going.

For me the stats that are important are:

Shooting percentage
Rebounding
Points in the paint.
Turnovers

Teams that are efficient win, and teams that own the paint win. The turnovers are thrown in, because they are an intangible that can still effect the game outside of those 3 above it.

 

[ABM]

To me it's meaningless.

As are most stats.

You'd never win-over Billy Bean or Theo Epstein.

Oh, WAIT! Different sport.

Nevermind.

 

[crowTrobot]

for the season it means outlaw sucks big time. it can't be easy to play more than half the time for a 21-14 team and still be -38

http://www.82games.com/0809/0809POR1.HTM

 

[Foulzilla]

Just like any other stat, it's useful in the right situation. There is no perfect stat. I do think +/- requires more scrutiny than most before drawing a conclusion from it, but it has it's uses. Primarily as a long term stat in my opinion, as a single game can be highly misleading.

 

[THE HCP]

No. You put a scrub in the game when it's a blowout, he could have a +20!

 

[Kaydow]

for the season it means outlaw sucks big time. it can't be easy to play more than half the time for a 21-14 team and still be -38

http://www.82games.com/0809/0809POR1.HTM

In this case, it's meaningful. Travis has played terrible perimeter defense this year (well, not that it has ever been great) and has left many a shooter wide open. It's part of the reason that Nate said he'll go back to the PF (the other reason is obvious - Frye and Diogu suck) Travis will forget to box out at PF, and he'll be overmatched by bigger stronger PF's at times but it's the lesser of two evils. His +/- will probably go up some.

 

[PapaG]

You'd never win-over Billy Bean or Theo Epstein.

Oh, WAIT! Different sport.

Nevermind.

Or Kevin Pritchard. The Blazers rely in part of a basketball version of sabermetrics when analyzing players.

 

[Tortimer]

Means almost nothing IMO.

 

[Minstrel]

+/- is a very good concept, but has clear flaws in practice. The concept is great because it is very bottom-line...there's no abstraction, it's all about pure impact on the floor, including defense and "intangibles." If it helps your team on the scoreboard, in any way, it should be reflected in the +/-.

The main flaws are:

-Over a small sample size (like a game, or a few games), luck can entirely wash out the meaning. If a player makes a number of great passes and his teammates miss a bunch of easy, open shots, the player will look worse in +/- than he really was. Over a large sample size, though, luck will largely even out.

-It doesn't account for who's on the floor with the player. If you are a backup, you're likely to have a lower +/- because you're part of the group that tends to do worse on the scoreboard than the starting unit. Even if you, personally, are a good player, your teammates are going to be worse, giving you a lower +/- than if you played with the starters. Playing as a starter on a great team, with a higher team point differential, you're certain to post a better +/- than if you play on a bad team that has a negative point differential, even though your ability level will be the same. Raw +/- doesn't account for teammates, but there are Adjusted +/- numbers that do.

So, I'd say that raw +/- is mostly useful for comparing starters to other starters on the same team and reserves to other reserves on the same team, and over large stretches of the season...not on a per-game basis.