Don't you mean the Autotrader +/- Stat? :)

Hmmm, I need to think about this some more, but I'm still inclined to think that it's a more meaningful individual stat for hockey. To qualify what I originally said (which was a bit of a no-brain throwaway comment), it's not meaningless for basketball, but I think it's less meaningful for a variety of reasons. The main one being that you have four separate lines in ice hockey, so the chances of you being in the game when a score occurs is pretty low, especially given that scoring is a relatively rare occurrence there. So if you build up a healthy +/- differential on a hockey team over the course of a season, I think that says more about your impact than it does to say that LeBron James is currently +506.

I don't quite understand why ELG's general criticism (in the other thread) about +/- and hockey wouldn't apply equally to basketball... maybe he can elaborate.

I guess the group stats they've got there are more interesting (although they're very much out of date according to the stat search available on that page), but there's a lot of noise involved when it comes to basketball, and your scoring fortunes are much more closely tied to the fortunes of your team than in hockey. Erik Burdon is probably +100 by now, but what does that mean other than he comes on in garbage time and hits a lot of 3s in blowout wins?

Anyway, just some thoughts to throw around... I'm happy to be shown wrong here!

Nothing against Burdon btw... I'm all for more court-time for the Burd-man!

+/- over a season makes more sense in hockey. But I like it in the NBA on a per game basis.

It would also be interesting to see the average +/- per minute to see which players make the biggest impact without a bias towards those that play big minutes. E.g: Who would have been the leading +/- player on the Celtics team last year? KG? Posey? PP?

Just saw on NBA.com that they have the +/- per minute but it doesn't seem to be sortable. Did note that KG had the best +/- on the Celtics last year tho.

It would also be good to take into account the average team differentials because those stats at the moment favour teams with good records.

Eg. Cleveland's team +/- per minute is about +0.21 while Memphis' is -0.14.

Mike Conley is -0.12 per minute and Andersen Varejao is +.21 per minute.

So while both players have significantly different +/- scores they both are about equal value to their team in terms of +/- differential.

It would be interesting to see all nba players sorted on that basis.

Just tangentalising for a minute: what if you could prescribe +/- ratings for contributors to online forums (such as this one)? Somehow work it out based on other forumgoers' responses to the particular forumgoer, pageviews and other criteria I'm not clever enough to work out?

I am sure thedoctor would have one of the highest ratings on here: whatever happened to him/her?

I've never used it but I instinctively like it (particularly on a per minute basis). Like all stats you need to eliminate the outliers (Burdon's career +/- goes up pretty dramatically when he comes on in garbage time when they are up 20 but how does it go in garbage time when they are down 20?).

Seems to me it reflects a player's contribution to the non-tangibles (or non-stats things) we find hard to track. You boxed out immaculately and set some big screens and as a consequence your team gave up no offensive boards but the ball didn't fall your way and you have a 0 for boards/points in the stats. Yet your team made a 5-point break in that period of play.

Meanwhile, you refuse to box out, crash the offensive boards (without running the offense) and get 4 boards and 4 points during the five minutes, but your team goes down by 5 in that period. I know which player I want playing for me.

I think a focus on this sort of stat ensures players are focused on positive team outcomes rather than individual outcomes (which quite often are unfortunately at odds with each other). Eg, the player at last year's nationals who went after every rebound to ensure she finished in the top rebounders - didn't box out, took some off team-mates (with the occasional spilled ball which went to the other team), achieved her goal, BUT the team finished 8th. Operation successful - the patient died.

This could also be known as the Mugsy Bogues stat. I remember Lute Olsen (after the 90 World's I think) talking about why he took Bogues away. He said at the selection camp, it didn't matter who they put him with, his team won. His stats weren't great, he just won a lot. They ended up putting him with their worst group, and he still won. Nothing that regular stats would show up, but a +/- makes it quite obvious.

I played with a number of perennial All-Stars whose stats were fantastic, but you knew they would never win a championship (and didn't) because nobody made them accountable for the teams outcomes - only their own.

Its not really that important, its just interesting. I guess for a low score game like hockey it would be good for a season, though. I dont know anything about hockey to really comment...

In basketball i suppose its good to see what bench players make an impact. But then a certain player who helped make a spark early on can also be on the receiving end of a run by the other team, so that can screw up the +/-. but if you're a coach and your guy is getting 15 min a game, and is always in a +, surely you would consider giving him more court time? I think it would be better suited to quarters, which im sure team statisticians are all over.

Well said 'ankles'. I think it is a measure of the intangible. Closest thing we have to quantiying players with the 'x' factor.

ankles, great post. The bit about outliers and my comment about it being useful in "broader" terms covers hereschenes Burdon argument, IMO. Obviously there are places where it's not a concrete stat (like I said, come on for the final seconds of a quarter and someone hits a freak three in a close game, and you might be -2 despite playing well and going down by a point) but as a complementary statistical factor, it has to be fairly valuable?

Was talking to someone about it the other day and they said that in the Breakers, Boucher tends to rank really high on this scale. Shows what he contributes besides stats to that team.

I think it would have to be far more useful than in a sport like ice hockey where a lucky or unlucky score can have a greater impact on your data.

You can look at LeBron's +506 and the team's record and at least take a guess that he's a valuable player. Tops their rankings and is an MVP front-runner - I don't think that's just convenient. He's had the biggest impact on his team's success.

It's probably a better indicator in a team season box or in an individual game than in league-wide rankings where obviously the strength of each team comes so heavily into play.

Here's an interesting one: similarish minutes, yet Nocioni from the Bulls is +69, Thomas is -164. Bogut leads the Bucks with +76 while Ridnour (similarish minutes) is second worst with -38. Hilario at the Nuggets +195, Kleiza -5!

Bo, I think thedoctor's been on holiday; he posted from India the other week.

What about the starter that gets left on the court to work with the rookies / bench in a blow out.

Going by this +/- stat you could say that the starter would lose out.

A few weeks ago when Cairns got some points together in the last few minutes off of rookie mistakes, turnovers, bad passing etc the one starter on the floor is hurt by being the guy left on the floor by the coach as the on floor support officer.

He has done nothing wrong personally but his stats are impacted by others mistakes.

Will probably balance out a bit over time.

It's a valuable stat, but needs to be compared with the team margin. If your team wins by 20 and someone has a +5 they probably haven't made a huge contribution. If your team loses by 20 and anyone has a +, they probably have been one of your better players per minute. A bench player that consistently gets +s may be worth a look for more time.

Dench playing for the 36ers would have had a shocking -ve. Whatever the reason, when he stepped on court the team went backwards.

I'm not so convinced it works well across games, but it gives a good ranking of your players within a game and particularly the long end of the bench.

I agree that it is quite a worthwhile stat. I think, particularly for defensive specialists/hustle style players who don't get a lot of stats (ie. Dillion Boucher for the Breakers) it can help to better appreciate their worth.

It's obviously not fool proof, but if you look at any team you will generally see that the better defenders are frequently the ones with high +/- ratings. I'm a Houston fan and I frequently notice that T-Mac, despite being considered one of the elite players in the NBA, is frequently one of the worst Rockets when it comes to +/-, while Rafer Alston, a career 35-39% shooter, usually has a very high plus minus rating, largely because of his high assist-to-turnover ratio, and stealing ability.

On a side-note, my dad was born in Canada and I'm a HUGE hockey fan, and I can tell you that +/- ratings are one of the most important stats in Ice Hockey, particularly for defencemen. And IMHO, as far as hockey goes, +/- is extremely accurate in assessing the intangibles of a hockey player, such as their defensive ability, positioning, awareness, etc, and those are all intangibles that play a role in BBall in one way or another.

Also, FWIW, I think its more valuable in Ice Hockey, because of the fact that scores are much less frequent. I can tell you for a fact that many of the GMs in the NHL look at +/- when it comes to signing/trading for defencemen before they look at their goals and assists. And certainly, even when it comes to forwards, you don't want a point-a-game player (which is really good, for those of you who don't know much about hockey) if their +/- rating is frequently say -10, -20 over the course of most seasons.