Why does anyone like RAPM?

[Doctor MJ]

If I'm using only one stat, I'd prefer a stat in the XRAPM family (such as RPM or now, RAPTOR). It does not make sense to ignore the box score.

Personally I'm not a fan of box priors, or any priors that don't stem from the broader implications of "impact". To me it dilutes the fundamental purpose of RAPM. Not to say box stats should be ignored (they shouldn't), but rather used to compliment/contextualize impact data.

That said, you're a thoughtful poster and I like to keep an open mind. What would you say is the reason for your preference?

I think you're missing part of what I'm saying.

I'm saying that if I didn't have access to all the information that every modern NBA fan has, then XRAPM would really add value over RAPM because it alone would be factoring in all the box score data.

But I'm not a proponent of XRAPM stats because I do have access to all of that, and I resent the fact that XRAPM stats have now basically taken over "the cutting edge". I use these stats because they are what's available and I can still do something with them, but I'd rather have really good reliable RAPM and other +/- stats.

[GeorgeMarcus]

If I'm using only one stat, I'd prefer a stat in the XRAPM family (such as RPM or now, RAPTOR). It does not make sense to ignore the box score.

Personally I'm not a fan of box priors, or any priors that don't stem from the broader implications of "impact". To me it dilutes the fundamental purpose of RAPM. Not to say box stats should be ignored (they shouldn't), but rather used to compliment/contextualize impact data.

That said, you're a thoughtful poster and I like to keep an open mind. What would you say is the reason for your preference?

I think you're missing part of what I'm saying.

I'm saying that if I didn't have access to all the information that every modern NBA fan has, then XRAPM would really add value over RAPM because it alone would be factoring in all the box score data.

But I'm not a proponent of XRAPM stats because I do have access to all of that, and I resent the fact that XRAPM stats have now basically taken over "the cutting edge". I use these stats because they are what's available and I can still do something with them, but I'd rather have really good reliable RAPM and other +/- stats.

Well put, I understand what you're saying now.

I don't have an advanced stat background so I approach these things conceptually and learn as I go.

[udfa]

Personally I'm not a fan of box priors, or any priors that don't stem from the broader implications of "impact". To me it dilutes the fundamental purpose of RAPM. Not to say box stats should be ignored (they shouldn't), but rather used to compliment/contextualize impact data.

That said, you're a thoughtful poster and I like to keep an open mind. What would you say is the reason for your preference?

I think you're missing part of what I'm saying.

I'm saying that if I didn't have access to all the information that every modern NBA fan has, then XRAPM would really add value over RAPM because it alone would be factoring in all the box score data.

But I'm not a proponent of XRAPM stats because I do have access to all of that, and I resent the fact that XRAPM stats have now basically taken over "the cutting edge". I use these stats because they are what's available and I can still do something with them, but I'd rather have really good reliable RAPM and other +/- stats.

Well put, I understand what you're saying now.

I don't have an advanced stat background so I approach these things conceptually and learn as I go.

That is me as well. I've come to appreciate advanced stats not because I'm a mathhead but because they have proven their value to me through experience. When I watch games and compare it to these stats, it makes a great deal of sense why particular players rate well or poorly. Plus they provide an empirical basis to inform discussions. I was very impressed when years ago I found BBRef and looked up what WS were. Not only was I impressed by the concept but I was very strongly persuaded of its validity by the fact that you can sum the WS of a roster and almost always get a result that is within 2 wins of their actual total and often within 1 win. I expect that other advanced stats can do as well, but because BBRef exists it was very easy to verify WS experimentally using their team pages.

[Doctor MJ]

Personally I'm not a fan of box priors, or any priors that don't stem from the broader implications of "impact". To me it dilutes the fundamental purpose of RAPM. Not to say box stats should be ignored (they shouldn't), but rather used to compliment/contextualize impact data.

That said, you're a thoughtful poster and I like to keep an open mind. What would you say is the reason for your preference?

I think you're missing part of what I'm saying.

I'm saying that if I didn't have access to all the information that every modern NBA fan has, then XRAPM would really add value over RAPM because it alone would be factoring in all the box score data.

But I'm not a proponent of XRAPM stats because I do have access to all of that, and I resent the fact that XRAPM stats have now basically taken over "the cutting edge". I use these stats because they are what's available and I can still do something with them, but I'd rather have really good reliable RAPM and other +/- stats.

Well put, I understand what you're saying now.

I don't have an advanced stat background so I approach these things conceptually and learn as I go.

Nothing wrong with that, appreciate you're willing to listen, particularly to someone railing against the dominant paradigm.

[Blackmill]

Two things I haven't seen mentioned:

1. In my experience, RAPM-type models tend to have many "solutions" with nearly the same error. For instance, you could probably find two models with nearly the same error, but the first says Player X is a +7, and the other says Player X is a +2. Big difference, same error.

2. A team is often more (or perhaps less) than the sum of its parts. Skill sets can be complementary or not. Most RAPM models just add the scores assigned to each player. Sometimes you can get away with modelling a complex system in a much simpler way (i.e. Taylor Series). I don't think that applies here given what people often want RAPM scores to represent.

[penbeast0]

Why does anyone like RAPM?

Yeah I know, it's not even really music. Just a bunch of guys cursing over background music they usually stole from other artists without paying the royalties.

[Jaivl]

Why does anyone like RAPM?

Yeah I know, it's not even really music. Just a bunch of guys cursing over background music they usually stole from other artists without paying the royalties.

You wouldn't dare!

[JordansBulls]

Spoiler:

I think because it speeds up your computer and who doesn't like a fast computer?

[Colbinii]

Spoiler:

I think because it speeds up your computer and who doesn't like a fast computer?

Especially when the woman is gone