PDA

View Full Version : Does +/- mean anything to you?



Nick O
12-28-2011, 03:46 PM
lol i was just wondering because ive never really thought of plus / minus as being an important stat in basketball and you never really hear anyone talk about it.. I just ask because ive been noticing its a recorded stat on the NBA.COM game trackers (companions) and i was wondering if there is really any sort of coalation someone might see between +/- and a solid player.. now obviously Lebron or Durant will have good ones so the example i noticed from yesterday is Jimmer Fridette had a.. meh game yesterday yet led his team in +/- ratio.. mean anything? thanks :P

Bruno
12-28-2011, 06:22 PM
i like lookin' at it. i think its interesting.

I like it especially for seeing which teammates click with each other. in pervious years Gasol and Bynum would have much higher +/- when Odom was on the floor, opposed to the other seven footer.

If i remember correctly, Bosh had a similar stat last year- he had a much better +/- when one of LBJ or Wade were on the bench?

I like it for teammate comparison stuff like that.

lavilevi23
12-28-2011, 06:26 PM
no

hugepatsfan
12-28-2011, 06:29 PM
I don't think it means much for one player. But I think it can be useful if used for groupings. Helps build rotations.

PhillyFaninLA
12-28-2011, 06:35 PM
I think the +/- is the one stat that can take into account your entire performance and intangibles but there are still other things that need to be looked at because a + 5 on one team could be a - 20 on a another.

But I am a fan of it.

Hawkeye15
12-28-2011, 06:35 PM
as stated above, its more valuable for finding rotations. Using it to evaluate an individual has to much roster support effect.

ManRam
12-28-2011, 06:39 PM
I think it's telling, it just depends on how you're using it to illustrate the point you're making....

For example, in game 1, the Magic offense clearly was better when Jameer was in as opposed to Duhon, and the +/- reflected that. However, it usually isn't that cut and dry. Like stated above, it is indeed best used to find and analyze rotations and groups of players...not so much individuals.

Cal827
12-28-2011, 06:44 PM
Not too much. I mean if the rest of your team can't guard then you suffer for it also... As mentioned better for rotations....

An example of a skewed result that I saw was last night. In Miami's win over Boston, at about the last 3 mins of the 3rd quarter, Bosh and Wade were both over +20, while Lebron was at -2.

Another example could be when Kobe violated the Raptors for 81. The defense was bad, but I don't think the point guard deserves a negative result, just cause Kobe went off on our SG/SF... you know except for the fact that he wasn't quintuple teamed, but that's the coach's fault.

PhillyFaninLA
12-28-2011, 06:53 PM
Not too much. I mean if the rest of your team can't guard then you suffer for it also... As mentioned better for rotations....

An example of a skewed result that I saw was last night. In Miami's win over Boston, at about the last 3 mins of the 3rd quarter, Bosh and Wade were both over +20, while Lebron was at -2.

Another example could be when Kobe violated the Raptors for 81. The defense was bad, but I don't think the point guard deserves a negative result, just cause Kobe went off on our SG/SF... you know except for the fact that he wasn't quintuple teamed, but that's the coach's fault.


The bolded happens and is an important point, as are those days where you completely shut down the other team. Value in the +/- comes in trends and consistency or lack there of not on a specific game or two.

Hawkeye15
12-28-2011, 06:56 PM
The bolded happens and is an important point, as are those days where you completely shut down the other team. Value in the +/- comes in trends and consistency or lack there of not on a specific game or two.

that is another important point. +/- is useless until you get an acceptable sample size.

Raph12
12-28-2011, 07:04 PM
Depends on the sample size...

NYKalltheway
12-28-2011, 07:30 PM
It's only good for the coach... Fans will never understand it and will call for a player's head if he's on - or not enough +

Chacarron
12-28-2011, 09:21 PM
I've used it before to show Monta Ellis' impact on the court. Let's just say it doesn't do him any good.

Chronz
12-29-2011, 02:20 AM
If it has done anything its given guys who usually dont post great individual stats to have some sort of quantifiable worth. If your a player whos teammates generally play better when your on the floor, thats as good as scoring/stopping the pts yourself.

Of course finding proof of these kind of players takes years of data, at which point it becomes a matter of you believing if that player is still at that level. A guy like Battier has had year of strong defensive +/- marks, and if you look at 6 year samples you find some of the best defenders. I dont think its a coincidence but I can understand why its not concrete.

Nick O
12-29-2011, 02:55 PM
If it has done anything its given guys who usually dont post great individual stats to have some sort of quantifiable worth. If your a player whos teammates generally play better when your on the floor, thats as good as scoring/stopping the pts yourself.

Of course finding proof of these kind of players takes years of data, at which point it becomes a matter of you believing if that player is still at that level. A guy like Battier has had year of strong defensive +/- marks, and if you look at 6 year samples you find some of the best defenders. I dont think its a coincidence but I can understand why its not concrete.

so ill use the example i had then.. if say through 2 games the guy with best +/- on the scramento kings is Jimmer Fridette hes played alot of minutes but we all know hes not a defender that would more than likeley be coincidence however if that trend lasted a full season would that add some sort of worth to him? .. i am a Jimmer fan :P thats the only reason ive used that example lol

ChiSox219
12-29-2011, 04:03 PM
Great insight in this thread. One thing to add, there are people that have techniques to adjust +/- in order to neutralize the effect of teammates and opponents. This is a more telling picture than just simple +/-.

Nick O
12-29-2011, 04:38 PM
is there anyway of finding say who would have had the best +/- overall last year.. just out of interest

ChiSox219
12-29-2011, 04:49 PM
is there anyway of finding say who would have had the best +/- overall last year.. just out of interest

http://www.nba.com/statistics/plusminus/plusminus_sort.jsp?pcomb=1&season=22010&split=9&team=

Mudvayne91
12-29-2011, 04:51 PM
As pointed out early, finding rotations and such. If you were to dig deeper with points in the paint, I'd think +/- would be effective judging bigs.

John Walls Era
12-29-2011, 04:57 PM
Yes if that player is +20 every game.

Dolfan305
12-29-2011, 07:04 PM
+/- is ok I guess if it's used for a group of 3 or so players in together, but for an individual player it's useless

Wade>You
12-29-2011, 07:05 PM
Blake Griffin last night: 28pts 11Reb, -37.

The rest of his team did nothing, yet the players had a +/- of -11 and lower.

:crazy:

Hawkeye15
12-29-2011, 07:13 PM
http://www.nba.com/statistics/plusminus/plusminus_sort.jsp?pcomb=1&season=22010&split=9&team=

and this is a perfect example of why that stat is nearly useless when using it on individuals. Of course players that get minutes from stacked teams will have higher +/- rates than players that may be much better players but are on bad teams.

ChiSox219
12-29-2011, 07:21 PM
and this is a perfect example of why that stat is nearly useless when using it on individuals. Of course players that get minutes from stacked teams will have higher +/- rates than players that may be much better players but are on bad teams.

I agree


What do you think of this list:

http://basketballvalue.com/topplayers.php?mode=summary&sortnumber=94&sortorder=DESC&year=2010-2011

nyanks79
12-29-2011, 07:26 PM
Like any stat it depends on how people use it and if they are able to put it in perspective. I think it could be valuable for role players, agiain if its in perspective. But I do think it shows how good a guy like Collison or Varajo can be for a good team.

Hawkeye15
12-29-2011, 07:28 PM
I agree


What do you think of this list:

http://basketballvalue.com/topplayers.php?mode=summary&sortnumber=94&sortorder=DESC&year=2010-2011

better, I actually have looked at that before. But I will always have an issue with +/- when it comes to evaluating players, because it doesn't take in account for example, that Jason Terry, or Manu (2 obvious ones), get a ton of minutes against back up players for example, or that Ron Artest, even though he sucks at this point, will always carry a very positive +/- rating because he is usually running in a unit that includes Kobe/Gasol/Bynum.

The only real reason to care about it, is as Chronz pointed out. Seeing possibly trends over long periods of times. The other way, again as Chronz pointed out, is to justify a player being on the floor a lot that may seem to some fans looking at his numbers, as having no business out there. In that case, you still need to look at the various units that are strongest on teams, and seeing which players are always there. For instance, Ricky Rubio's +/- numbers in Europe over his pro career have been extremely good. His teams have just played better with him on the floor, even though many fans struggle to find his statistical contributions.

Again, its nice over the course of the season, and tells us which rotations are best when it comes to team play. But I could almost care less about what it tells me individually. Ron Artest is not better than Kevin Love, but according to +/- he sure is.

ChiSox219
12-29-2011, 07:35 PM
better, I actually have looked at that before. But I will always have an issue with +/- when it comes to evaluating players, because it doesn't take in account for example, that Jason Terry, or Manu (2 obvious ones), get a ton of minutes against back up players for example, or that Ron Artest, even though he sucks at this point, will always carry a very positive +/- rating because he is usually running in a unit that includes Kobe/Gasol/Bynum.

The only real reason to care about it, is as Chronz pointed out. Seeing possibly trends over long periods of times. The other way, again as Chronz pointed out, is to justify a player being on the floor a lot that may seem to some fans looking at his numbers, as having no business out there. In that case, you still need to look at the various units that are strongest on teams, and seeing which players are always there. For instance, Ricky Rubio's +/- numbers in Europe over his pro career have been extremely good. His teams have just played better with him on the floor, even though many fans struggle to find his statistical contributions.

Again, its nice over the course of the season, and tells us which rotations are best when it comes to team play. But I could almost care less about what it tells me individually. Ron Artest is not better than Kevin Love, but according to +/- he sure is.

I look at it as a tool to measure performance in a given role.

Hawkeye15
12-29-2011, 07:39 PM
I look at it as a tool to measure performance in a given role.

I can see that as well. Do you ever look at 82 games.com's most productive units, and what a player does at various positions? I am trying to buy into that some, but it still has some limiting factors that won't allow me to just use those numbers as evidence, as far as the positional PER's a player produces and gives up. For instance, Love allows a PER of like 12.7 to a Center, but I think it has more to do with they only play Love at center when the other team has a weak center lineup.

PatsSoxKnicks
12-29-2011, 08:10 PM
I agree


What do you think of this list:

http://basketballvalue.com/topplayers.php?mode=summary&sortnumber=94&sortorder=DESC&year=2010-2011

RAPM is probably better.

http://stats-for-the-nba.appspot.com/ranking11

I still need to spend some time trying to figure out how RAPM works. I did do that with adjusted +/-, which I like the concept of but it still needs more years of data before it becomes really effective.

I mean the standard errors are around 5 for 1 year APM and Rose's standard error last year was around 8.75. So his +/- numbers could've been anywhere from 3 to 20. That's a very large range. Even 2 year APM has a standard error from about 2.5-4, which is still decently large. I think it was found that 6 year APM was the best for minimizing the standard error but then the problem you run into is whether the player stays the same over those 6 years.

RAPM apparently solves some of these issues.

Hawkeye15
12-29-2011, 08:13 PM
RAPM is probably better.

http://stats-for-the-nba.appspot.com/ranking11

Is it wrong that I look for Love immediately, and if I see him that far down, I disregard the list haha?

PatsSoxKnicks
12-29-2011, 08:24 PM
Is it wrong that I look for Love immediately, and if I see him that far down, I disregard the list haha?

lol, the list does seem a bit weird. I know that Jeremy Engleman is the guy who runs that site but the man who actually implemented the idea was Joe Sill, so there could be differences between how Sill calculates it and how that site does it.

In any case, whenever I get the chance to meet with Oliver, I plan on asking him some questions about RAPM. I'm still not sure how valid it is (well I know Oliver believes in it). I've also been told its easy to understand so maybe I just need to read Sill's paper.

Hawkeye15
12-29-2011, 08:30 PM
lol, the list does seem a bit weird. I know that Jeremy Engleman is the guy who runs that site but the man who actually implemented the idea was Joe Sill, so there could be differences between how Sill calculates it and how that site does it.

In any case, whenever I get the chance to meet with Oliver, I plan on asking him some questions about RAPM. I'm still not sure how valid it is (well I know Oliver believes in it). I've also been told its easy to understand so maybe I just need to read Sill's paper.

Ask him to give Love some ****ing props!

My worst fear is coming true. I am slowly morphing into a homer in regards to a single player on my team...

Bruno
12-29-2011, 08:41 PM
If it has done anything its given guys who usually dont post great individual stats to have some sort of quantifiable worth. If your a player whos teammates generally play better when your on the floor, thats as good as scoring/stopping the pts yourself.

Of course finding proof of these kind of players takes years of data, at which point it becomes a matter of you believing if that player is still at that level. A guy like Battier has had year of strong defensive +/- marks, and if you look at 6 year samples you find some of the best defenders. I dont think its a coincidence but I can understand why its not concrete.

Great article on Battier, it really expands on what Chronz is talking about here:

http://www.nytimes.com/2009/02/15/magazine/15Battier-t.html?pagewanted=all



What do you think of this list:

http://basketballvalue.com/topplayers.php?mode=summary&sortnumber=94&sortorder=DESC&year=2010-2011

I thought that was pretty cool.


Ask him to give Love some ****ing props!

My worst fear is coming true. I am slowly morphing into a homer in regards to a single player on my team...

haha. that'll happen.

Nick O
12-29-2011, 08:51 PM
I agree


What do you think of this list:

http://basketballvalue.com/topplayers.php?mode=summary&sortnumber=94&sortorder=DESC&year=2010-2011

thanks :) i find that one more helpful

Tony_Starks
12-30-2011, 03:45 AM
Other than helping psd people win arguments, nope not really.....

Chronz
12-30-2011, 04:17 AM
so ill use the example i had then.. if say through 2 games the guy with best +/- on the scramento kings is Jimmer Fridette hes played alot of minutes but we all know hes not a defender that would more than likeley be coincidence however if that trend lasted a full season would that add some sort of worth to him? .. i am a Jimmer fan :P thats the only reason ive used that example lol

I dont know what the official bench mark is but the greater the sample size the better. 1 year samples can still be flukish IMO as far as I can tell 2 years seems to be the middle ground people look at, since we have data dating from 2003 I just look at the #'s in their entirety from that starting point, if I spot some consistent trend in any of the various forms of +/- then I feel comfortable citing the players stats in that way.

Its useful in finding value in players (by use of stats) you would otherwise have no idea they contributed. Like Jason Collins back in the early 2000's, hes probably one of the worst players of all time statistically, but his defensive markers were always decent according to +/-, his teams always defended and rebounded better with him on the floor despite his middling rebounding/shot blocking averages. Then +/- got more advanced and you were able to look at the stats a player put up when a certain player was on the floor, and as Jason Kidd will tell you, Collins boxing out skill is a big reason why he was able to corral so many boards. So many of Collins teammates rebounded better with him on the floor that, over time, it became obvious it wasnt a fluke. This doesnt completely exonerate Collins, had he been a better rebounder himself, his team would have been dominant on that end but if you can keep your conterparts rebounds down, these stats make up for your own lack of rebounding in some way.

Chronz
12-30-2011, 04:18 AM
Other than helping psd people win arguments, nope not really.....

You mean other than helping the Mavs win titles, nothing really