RealGM

How do you guys feel about RAPM? It took me a long time to warm up to this stat. At first I was very against it, mainly because I didn't understand the math behind it (and still don't) and it did have a few funky readings. After time, the more and more I looked into RAPM the more and more it seemed to line up with the eye test and other statistics. There are a few anomalies that tend to arise but if you examine the context around those certain higher than expected scoring players, (or lower than expected scoring players), you can usually make sense out of it.

I personally don't know much about it but just like any other advanced stat, it should be used as tool and not the only thing you should use.

It's the best stat currently in existence.

Some people will scoff at this, but just think on this: what are we actually trying to evaluate when we look at a basketball player? Answer: How well does he help his team win? (Can be restated as: what effect does he have on the scoring margin?) The reason I hold RAPM in such high regard is because it is literally the only stat that actually attempts to answer this question. Any box score stat you can think of doesn't even try. Quite literally, RAPM is the only stat that has any validity for what people are actually looking for in a stat, even if they don't quite realize it.

There's nothing inherently wrong with using the box score, as long as you realize that it is, at best, a proxy for what you actually want to know. There can be great players who score 20+ ppg, and terrible ones who do the same. But there will never ever, by definition, be a terrible player who makes a hugely positive impact on his team.

Once I came to realize this, I became a big RAPM convert, and I live with the flaws because it's the only thing that can actually tell me what I want to know. It might miss the mark by more than the box score will, but at least the mark in this case is clear, and it's exactly what I want it to be.

So really this is the whole reliability vs. validity issue. People dislike RAPM generally because it challenges what they thought they knew. The fallacy is in thinking that you ever knew anything by looking at the box score anyway. So yeah, in small samples we have some results that are absolutely nuts. Granted. But with big samples, and enough noise correction, we zero in on exactly what we really want to know. That's beautiful, and it's something no other stat in existence can accomplish in the slightest.

Now I'll add the caveat that RAPM never thinks for me. I would never use it to rank players, or the crux of an argument, or anything of the sort. But as you guys know, I watch a ton of film, and generally find that what I see lines up with what plus/minus data shows. Obviously Manu Ginobili isn't the best player in the league, but when he comes in the Spurs play really well, and that's what RAPM tells us. Now Kevin Durant has a lower RAPM, but a much huger role and more minutes, so quite obviously he's the better player. I think sometimes people who use RAPM get pisgeonholed into using it as the be-all-end-all, and that's not what I do at all.

Final thing: RAPM is at a crossroads right now, and it's either going to head into RPM (wrong direction) or PTPM (right direction). This has everything to do with how I feel about the box score.

^
A SPM model does try to answer the same question.

Generally speaking, the people who grok the math behind RAPM see its merits while recognizing it's not the holy grail, final word authority.

Of the posters who are most vehemently opposed, I haven't seen anyone demonstrate even a rudimentary knowledge of linear algebra. Their objections are almost always centered around results that are contrary to what they already believe.

Has to be the best defensive stat, there isn't much competition there, plus defensive value just seems more consistent from team to team and lineup to lineup. Offense I think runs into more "correlates with success= not as much the cause of its success as RAPM says" issues

No problem with the stat at all. It's clearly useful information. My only issue with it is when guys start there and work backwards trying to explain why a player beloved by RAPM is so great instead of using as one tool among many to try and gauge a player's true impact. But that's not a complaint about the stat, but rather how it gets used incorrectly. Unforunately too many guys hear its the best stat and assume it is isolating individual impact perfectly.

I don't like it, because there's no way to break it down and analyze where the numbers came from. You're just given a number and have to take it on faith that it's accurate, even though it's based on a methodology that will obviously produce innaccuracies (trying to control for performance of the other players on the court, when those performances can vary a lot from game to game). At least with box score numbers, you know that the numbers are accurate and you know where they came from, and it becomes a matter of analyzing them in context and coming to conlusions about how impressive they actually are and what they indicate about a player's value.

Dr Spaceman wrote:It's the best stat currently in existence.

Some people will scoff at this, but just think on this: what are we actually trying to evaluate when we look at a basketball player? Answer: How well does he help his team win? (Can be restated as: what effect does he have on the scoring margin?) The reason I hold RAPM in such high regard is because it is literally the only stat that actually attempts to answer this question. Any box score stat you can think of doesn't even try. Quite literally, RAPM is the only stat that has any validity for what people are actually looking for in a stat, even if they don't quite realize it.

There's nothing inherently wrong with using the box score, as long as you realize that it is, at best, a proxy for what you actually want to know. There can be great players who score 20+ ppg, and terrible ones who do the same. But there will never ever, by definition, be a terrible player who makes a hugely positive impact on his team.

Once I came to realize this, I became a big RAPM convert, and I live with the flaws because it's the only thing that can actually tell me what I want to know. It might miss the mark by more than the box score will, but at least the mark in this case is clear, and it's exactly what I want it to be.

So really this is the whole reliability vs. validity issue. People dislike RAPM generally because it challenges what they thought they knew. The fallacy is in thinking that you ever knew anything by looking at the box score anyway. So yeah, in small samples we have some results that are absolutely nuts. Granted. But with big samples, and enough noise correction, we zero in on exactly what we really want to know. That's beautiful, and it's something no other stat in existence can accomplish in the slightest.

Now I'll add the caveat that RAPM never thinks for me. I would never use it to rank players, or the crux of an argument, or anything of the sort. But as you guys know, I watch a ton of film, and generally find that what I see lines up with what plus/minus data shows. Obviously Manu Ginobili isn't the best player in the league, but when he comes in the Spurs play really well, and that's what RAPM tells us. Now Kevin Durant has a lower RAPM, but a much huger role and more minutes, so quite obviously he's the better player. I think sometimes people who use RAPM get pisgeonholed into using it as the be-all-end-all, and that's not what I do at all.

Final thing: RAPM is at a crossroads right now, and it's either going to head into RPM (wrong direction) or PTPM (right direction). This has everything to do with how I feel about the box score.

Can you name five "terrible" players who score 20+ ppg?

Agree with Chuck. Its useful, but definitely has limits to it as well. I think it conveys good information but can't take context into account, which is a big issue given that sometimes context can be a huge factor in overall evaluation. If you're using it as the ultimate player ranking, then I'd disagree, but if you're checking into it as something to go along w/ other stats and watching the games, its useful.

It's a much better advanced stat than any box score based stat (PER, BPM, etc.), but it's still only really useful for players with 3+ years played and even then, only when put into proper context.

Keller61 wrote:Can you name five "terrible" players who score 20+ ppg?

Not sure if "terrible", but pretty meh:

*Pre-Grizzlies Zach Randolph
*Al Jefferson
*Jerry Stackhouse
*Purvis Short
*Rolando Blackman

Jaivl wrote:

Keller61 wrote:Can you name five "terrible" players who score 20+ ppg?

Not sure if "terrible", but pretty meh:

*Pre-Grizzlies Zach Randolph
*Al Jefferson
*Jerry Stackhouse
*Purvis Short
*Rolando Blackman

To add some more:
Lebron, KD, and Melo in their rookie years
Antoine Walker
Andrea Bargnani
Mike James

I don't like it one bit.

I think it's forced a lot of posters into reverse-logic: take something you know to be true and change those beliefs because of RAPM. I also hate people just discounting the noise as it if doesn't mean anything.

I dislike it intensely and will never use it as a part of my analysis. Worsens the discussion around here a good bit, and ruined the top 100 project for me.

Dr Spaceman wrote:It's the best stat currently in existence.

Some people will scoff at this, but just think on this: what are we actually trying to evaluate when we look at a basketball player? Answer: How well does he help his team win? (Can be restated as: what effect does he have on the scoring margin?) The reason I hold RAPM in such high regard is because it is literally the only stat that actually attempts to answer this question. Any box score stat you can think of doesn't even try. Quite literally, RAPM is the only stat that has any validity for what people are actually looking for in a stat, even if they don't quite realize it.

Once I came to realize this, I became a big RAPM convert, and I live with the flaws because it's the only thing that can actually tell me what I want to know. It might miss the mark by more than the box score will, but at least the mark in this case is clear, and it's exactly what I want it to be.

So really this is the whole reliability vs. validity issue. People dislike RAPM generally because it challenges what they thought they knew. The fallacy is in thinking that you ever knew anything by looking at the box score anyway. So yeah, in small samples we have some results that are absolutely nuts. Granted. But with big samples, and enough noise correction, we zero in on exactly what we really want to know. That's beautiful, and it's something no other stat in existence can accomplish in the slightest.

Look, you're usually a very good poster but I hope this is sarcasm cus you make it sound like RAPM is some sort of religion with the way you pour out unconditional confidence in it.

The box score is still the most important statistic. Minutes played, FGA, Assists, etc...without them as a background on a player's role, RAPM becomes useless.

Swagalicious wrote:

Dr Spaceman wrote:It's the best stat currently in existence.

Some people will scoff at this, but just think on this: what are we actually trying to evaluate when we look at a basketball player? Answer: How well does he help his team win? (Can be restated as: what effect does he have on the scoring margin?) The reason I hold RAPM in such high regard is because it is literally the only stat that actually attempts to answer this question. Any box score stat you can think of doesn't even try. Quite literally, RAPM is the only stat that has any validity for what people are actually looking for in a stat, even if they don't quite realize it.

Once I came to realize this, I became a big RAPM convert, and I live with the flaws because it's the only thing that can actually tell me what I want to know. It might miss the mark by more than the box score will, but at least the mark in this case is clear, and it's exactly what I want it to be.

So really this is the whole reliability vs. validity issue. People dislike RAPM generally because it challenges what they thought they knew. The fallacy is in thinking that you ever knew anything by looking at the box score anyway. So yeah, in small samples we have some results that are absolutely nuts. Granted. But with big samples, and enough noise correction, we zero in on exactly what we really want to know. That's beautiful, and it's something no other stat in existence can accomplish in the slightest.

Look, you're usually a very good poster but I hope this is sarcasm cus you make it sound like RAPM is some sort of religion with the way you pour out unconditional confidence in it.

The box score is still the most important statistic. Minutes played, FGA, Assists, etc...without them as a background on a player's role, RAPM becomes useless.

Yeah. I'd rather look at box score numbers and results and form my own conclusions than use any kind of all-in-one stat like RAPM or PER.

Dr Spaceman wrote:It's the best stat currently in existence.

Some people will scoff at this, but just think on this: what are we actually trying to evaluate when we look at a basketball player? Answer: How well does he help his team win? (Can be restated as: what effect does he have on the scoring margin?) The reason I hold RAPM in such high regard is because it is literally the only stat that actually attempts to answer this question. Any box score stat you can think of doesn't even try. Quite literally, RAPM is the only stat that has any validity for what people are actually looking for in a stat, even if they don't quite realize it.

There's nothing inherently wrong with using the box score, as long as you realize that it is, at best, a proxy for what you actually want to know. There can be great players who score 20+ ppg, and terrible ones who do the same. But there will never ever, by definition, be a terrible player who makes a hugely positive impact on his team.

Once I came to realize this, I became a big RAPM convert, and I live with the flaws because it's the only thing that can actually tell me what I want to know. It might miss the mark by more than the box score will, but at least the mark in this case is clear, and it's exactly what I want it to be.

So really this is the whole reliability vs. validity issue. People dislike RAPM generally because it challenges what they thought they knew. The fallacy is in thinking that you ever knew anything by looking at the box score anyway. So yeah, in small samples we have some results that are absolutely nuts. Granted. But with big samples, and enough noise correction, we zero in on exactly what we really want to know. That's beautiful, and it's something no other stat in existence can accomplish in the slightest.

Now I'll add the caveat that RAPM never thinks for me. I would never use it to rank players, or the crux of an argument, or anything of the sort. But as you guys know, I watch a ton of film, and generally find that what I see lines up with what plus/minus data shows. Obviously Manu Ginobili isn't the best player in the league, but when he comes in the Spurs play really well, and that's what RAPM tells us. Now Kevin Durant has a lower RAPM, but a much huger role and more minutes, so quite obviously he's the better player. I think sometimes people who use RAPM get pisgeonholed into using it as the be-all-end-all, and that's not what I do at all.

Final thing: RAPM is at a crossroads right now, and it's either going to head into RPM (wrong direction) or PTPM (right direction). This has everything to do with how I feel about the box score.

Basically what you're saying in all of this is, "RAPM is better than all other measures because i said so, and anyone who doesn't use RAPM is wrong or ignorant."

I don't know what to make of it. I've only really tried following it since I came here in the last year. Obviously the behind-the-scenes number crunching where consistency depends on the model chosen is problematic on the face. Beyond that it seems to attempt to answer a really specific question, the utility of which I'm not exactly positive translates into player-to-player comparisons much. I don't discount the power of formulas, merely the extent of it's reach.

I do understand the appeal. We can't watch all games, so best to have a stat that we can plug all games into and tell us what we missed. This may be a step in that direction, but it isn't that. And so it remains a curiosity for me, something I might look at within limits.

Chuck Texas wrote:No problem with the stat at all. It's clearly useful information. My only issue with it is when guys start there and work backwards trying to explain why a player beloved by RAPM is so great instead of using as one tool among many to try and gauge a player's true impact. But that's not a complaint about the stat, but rather how it gets used incorrectly. Unforunately too many guys hear its the best stat and assume it is isolating individual impact perfectly.

bondom34 wrote:Agree with Chuck. Its useful, but definitely has limits to it as well. I think it conveys good information but can't take context into account, which is a big issue given that sometimes context can be a huge factor in overall evaluation. If you're using it as the ultimate player ranking, then I'd disagree, but if you're checking into it as something to go along w/ other stats and watching the games, its useful.

Echoing these sentiments. I consistently looked at RAPM data (and the like) during the top 100 project, but stayed away from using it in my arguments as I just didn't feel comfortable with it yet. I'm not against it by any means as any data we can add to the table and look at when evaluating players, teams, etc. is welcomed.

I just don't agree with using it as the definitive measure for players on its own. I certainly believe in impact beyond the box score, though.

Swagalicious wrote:Look, you're usually a very good poster but I hope this is sarcasm cus you make it sound like RAPM is some sort of religion with the way you pour out unconditional confidence in it.

I was over-the-top hyperbolic, but I think I got my point across. The main thing is, if you're not a believer in RAPM, what are you putting faith in? People just accept the box score without really thinking about what it is or what it measures, and just sort of assume it's okay to tell us who is having positive impact and who's not. RAPM has problems, but it does a better job of this than the box score.

Swagalicious wrote:The box score is still the most important statistic. Minutes played, FGA, Assists, etc...without them as a background on a player's role, RAPM becomes useless.

Those things are more important than team performance? Being a little less withholding: I want to make clear that I don't hold RAPM as gospel in the slightest. I just think it happens to be way more useful than anything else out there.

mischievous wrote:Basically what you're saying in all of this is, "RAPM is better than all other measures because i said so, and anyone who doesn't use RAPM is wrong or ignorant."

[/quote]

No, that's not at all what I'm saying. I'm saying that if you're going to use statistics, RAPM is the one that has the most validity.