Engelmann Playoff only RAPM (1997-2024)

[lessthanjake]

.

I agree completely, noisy samples like this or like the great work squared and others did with pre 2000's rapm are nice to have but extremely flawed to actually use in comparision to bigger and complete data sets of other players (or the same player*)

Yep—sample-size problems exist for Squared’s RAPM too. Sample-size issues are inherently a matter of degree though, and at this point with Squared’s RAPM the samples are generally bigger than these playoff samples are. Squared’s RAPM also doesn’t have the other issue I described, where “off” samples for superstars in the playoffs tend to be more consumed by garbage time. Not to mention that the playoff “off” sample will tend to be noticeably smaller for any given number of games played, because stars play more minutes per game in the playoffs. Because of these factors, the noise in a playoff sample is very likely going to be significantly higher than the noise in the same number of games in the regular season. When combined with the fact that the number of games in the overall Squared sample is at the point where it’s generally larger than these playoff samples, and I think there’s reason to think that sample-size issues are a noticeably smaller concern there. But it is certainly still a concern.

Of course, sample-size issues are not the only issue with Squared’s RAPM. Notably, because it does not include all games, Squared’s RAPM also has potential sampling error that isn’t here for most players in the data set (there is sampling error for players who played playoff games before 1997, but that’s not most players in the 1997-2024 span). So there’s good reason to take Squared’s data with a grain of salt, even if the sample-size issue isn’t as much of a problem as it is here.

The other thing to emphasize with the limitations with playoff-only RAPM is just how limiting the lack of opponent diversity is.

Players face -- at maximum -- four different teams each season. That's miniscule. Even across careers, that's miniscule. Conferences face each other *once* in playoff-only samples, and those are only from a single team. The only plus minus signal we get to scale eastern conference vs western conference players are the finals, and the times players switch conferences. Again, minuscule, and highly limiting.... and completely different from the Squared2020 data we have. These kind of limitations to the number of opposing lineups makes accurate and unbiased RAPM calculations vastly more difficult, and again it's far more pervasive to this playoff-only data than the historical data.

This is a good point about playoff RAPM, which I’d not previously thought about. I will say that Squared’s data does still have a bit of an issue that’s at least of the same genre, since there’s some teams that are not sampled much at all, while others are sampled a lot. The Squared data is definitely tilted towards a specific set of teams and opponents. Not to the same degree as playoff data inherently is, of course, but it’s not got perfect opponent diversity either.

[DraymondGold]

Yep—sample-size problems exist for Squared’s RAPM too. Sample-size issues are inherently a matter of degree though, and at this point with Squared’s RAPM the samples are generally bigger than these playoff samples are. Squared’s RAPM also doesn’t have the other issue I described, where “off” samples for superstars in the playoffs tend to be more consumed by garbage time. Not to mention that the playoff “off” sample will tend to be noticeably smaller for any given number of games played, because stars play more minutes per game in the playoffs. Because of these factors, the noise in a playoff sample is very likely going to be significantly higher than the noise in the same number of games in the regular season. When combined with the fact that the number of games in the overall Squared sample is at the point where it’s generally larger than these playoff samples, and I think there’s reason to think that sample-size issues are a noticeably smaller concern there. But it is certainly still a concern.

Of course, sample-size issues are not the only issue with Squared’s RAPM. Notably, because it does not include all games, Squared’s RAPM also has potential sampling error that isn’t here for most players in the data set (there is sampling error for players who played playoff games before 1997, but that’s not most players in the 1997-2024 span). So there’s good reason to take Squared’s data with a grain of salt, even if the sample-size issue isn’t as much of a problem as it is here.

The other thing to emphasize with the limitations with playoff-only RAPM is just how limiting the lack of opponent diversity is.

Players face -- at maximum -- four different teams each season. That's miniscule. Even across careers, that's miniscule. Conferences face each other *once* in playoff-only samples, and those are only from a single team. The only plus minus signal we get to scale eastern conference vs western conference players are the finals, and the times players switch conferences. Again, minuscule, and highly limiting.... and completely different from the Squared2020 data we have. These kind of limitations to the number of opposing lineups makes accurate and unbiased RAPM calculations vastly more difficult, and again it's far more pervasive to this playoff-only data than the historical data.

This is a good point about playoff RAPM, which I’d not previously thought about. I will say that Squared’s data does still have an bit of an issue that’s at least of the same genre, since there’s some teams that are not sampled much at all, while others are sampled a lot. The Squared data is definitely tilted towards a specific set of teams and opponents. Not to the same degree as playoff data inherently is, of course, but it’s not got perfect opponent diversity either.

Oh of course! It’s a partial sample and far from perfect.

But like you say, there is significantly more mixing of opponents and conferences in the historical data than the playoff only data. Even with the partial sample, I would bet there’s easily at least over an order of magnitude more mixing of opponents in the historical data than the playoff data.


There’s definitely a selection bias towards games from certain teams like you say. But within a team, the games aren’t intentionally biased high or low — it’s just whatever games are available. And unlike the data here, Squared provides both the sampling bias compared to the full-season data and the uncertainty bars, which are both key to interpretation.

Again none of this is to rag on Engelmann. It’s really interesting to see the playoff only RAPM, and I appreciate that they make all this public. I’ve been posting and interpreting the data basically every time there’s an Engelmann career RAPM thread (while also emphasizing the limitations each time, like you). I’ll keep posting in each of these threads, because I find the results genuinely interesting! But there’s clear reason to prefer the RS-only or full-season or age-curve-corrected RAPM to the PS-only RAPM here.

[homecourtloss]

it's always interesting that when there is some partial RAPM data released by squared or partial on/off data calculated by somebody or released by squared or somebody else and even if you have incomplete data or small off samples or whatever, as long as those data sets show certain players looking good there is no "cautioning," no "not very useful," "too small amount of data," for the various other caveats that you see not only in this thread, but every thread. After the usual "cautioning" and cook take it with a grain of salt" and "small/tiny off samples," you see long posts talking about why the data isn't useful or not to be taken too seriously even though those very same cautioning or precautions aren't done with other data that also has small and/or incomplete sample sizes. In fact, those are the threats are responded to with "this is amazing work!" "this dara really shows something."

I am 100% sure that had this particular set not had LeBron far away at the top or curry below Draymond or Jokić nowhere near the top which is reflective of his playoffs on/off, etc., these very same cautioning, but of course not be given.

Consistency would be nice.

I’m sorry but this is a bit maddening. It seems fairly clear that you’re at least in part meaning to refer to me, since your post comes after a post that responded to me about this same issue (not to mention that you frequently make snide remarks directed at me, and your post’s other examples of Curry/Draymond and Jokic completely align with this referring to me). But you’ll find that I’m actually really consistent in caveating and cautioning regarding Squared’s data. I can’t even count how many times I’ve done so. Indeed, I actually kind of doubt there’s anyone on these forums who has done so more times than I have. I even just did it in response to Falco’s post, before I saw your post (ironically, I even used some of the exact same language you claim is never used regarding Squared, specifically saying it should be taken with “a grain of salt”)! The idea that “those same cautioning and precautions aren’t done with other data” is just manifestly false if you’re referring to me (which you definitely are). You’ll also find that I’m very consistent about expressing concerns regarding playoff RAPM, not just in this thread. Again, the idea that my concerns in this arena are motivated by the results of any particular measure is nonsense that is completely divorced from my actual post history. It’s just a complete straw man, and I find it maddening because it’s not the first time you’ve made posts like this. You tend to make posts where you are obviously talking about me without explicitly saying you’re talking about me, while making assertions/accusations about my post history that are just objectively untrue. I’d appreciate you not doing so. I can understand that you probably have not have read all my posts and that that could potentially be a good faith cause of your incorrect impressions, but even if that’s the case I’d say that you shouldn’t make broad assertions/accusations about my posts if you haven’t actually done your homework enough to know that what you’re saying is accurate.

The maddening thing is when posters feign neutrality sans agenda when what they post is very often that. I think it's ok to be partisan and out of partisanship we might collect/post new and interesting data like a Cold Era arms race leading to greater tech. Claim it and own it rather than hide behind some sort of objectivity, "i'm being mischaracterized," etc.

This is your first post right after the data release:

Given how tiny playoff off samples are, I think this is mostly extremely noisy to the point of being not very useful, except perhaps to look at the small number of people with tons of playoff games in this data set (guys like LeBron, Steph, Draymond, Duncan, and Ginobili), and even some of those wouldn’t be much less noisy than single-season RAPM (which is super noisy). And it’s actually probably even worse than just looking at the off minutes would lead you to believe, because, even for those players with lots of playoff minutes, the model will be trying to control for the presence of other players that have a very small amount of playoff data. So, basically, the adjustments the model makes compared to on-off will be mostly based on garbage.

There is a straightaway dismissal of the data. There's lots of other data posted that's incomplete, noisy, but for that data, there isn't a straightaway dismissal. I wonder why.

Had this had a different player at the top or if Jokic looked better it Curry over Draymond, etc., I'm confident that the first post would not be a straightaway dismissal. Maybe we'd get



Or from the other poster here feigning the same we'd get



or

instead of a 1,000 word dismissal.

Again, consistency and owning it.

[f4p]

Really interesting to think about what the implications of this are for Draymond. I kinda had an inkling he was more crucial to the Warriors than regular season impact numbers would suggest. And now that I think about it, his defense always was a lot more valuable in the playoffs than the regular season. I think I’m going to have start thinking seriously about a top 30 all-time case for him.

to me, the biggest drawback is that, just like steph, it's not clear either have the same value without each other. their combined lineups have always been ridiculous, because they are probably the two most synergistic [high level] teammates ever. like usually you would think with player A and player B, their value would be close to:

RAPM A + RAPM B

but for steph and draymond it feels like it's more like

RAPM A + RAPM B + 2 or 3 more points

which makes evaluating both of them difficult, because i don't think either is as valuable in another situation. and i think it applies more to draymond obviously, as steph's skills are fairly transferable, even if he doesn't ever find his draymond, but even steph had a perfect system built around him by a perfect coach with a perfect teammate like draymond to help run the system (and also iggy) without steph having to be on ball like he probably would be in a lot of his potential careers. as for draymond, i can buy draymond as top 30 for what he's done on the warriors and the massive impact that comes with it, but that feels like a 25% case for his career (or probably less). averaged over all his potential careers, i think his 50 or so ranking in the last Top 100 feels more fair.

It's seriously impressive that a Draymond-reliant Warriors could perform well in the first round without Curry.

it's not just impressive, it was necessary. steph has only played 7 healthy playoffs in his career, way less than most any other superstar. without draymond wrecking 1st round teams, steph missed out on a finals in 2016 and a championship in 2018. a lot different resume without that.

But I don't think this is compelling evidence to think Draymond was more valuable to the playoff Warriors than Curry was over the course of their careers. I think the sample is just being biased here by the easier first-round opponents and the easier schedule that Draymond faced, which is exactly the kind of thing a playoff-only RAPM would miss if it ignores all regular season data.

why isn't it compelling evidence? after all, RAPM either works or it doesn't, right? do i personally believe he was "more" valuable? probably not. though it seems unlikely that there is much difference. but like i said below, i don't weight RAPM as much as others. for those that do, this is exactly what the data is telling us. after all, isn't RAPM supposed to be able to account for opponent strength and adjust for it? so a first round opponent being bludgeoned or a close win over a great team should look similar. we can't just pick the samples where draymond only looks equal instead of better.

[f4p]

OK, based on this I’m immediately deciding that I’ve underrated Kevin Durant.

yeah i've been told KD is a value-less hack who doesn't provide defensive value and can't really run an offense and that he's just a throw in on teams. weird that RAPM doesn't feel the same way. also, i was told james harden's isoball didn't work in the playoffs.

of course, i'm not as big of a believer in RAPM as many other, but for the ones who are, it feels like this should move rankings.

[f4p]

Given how tiny playoff off samples are, I think this is mostly extremely noisy to the point of being not very useful, except perhaps to look at the small number of people with tons of playoff games in this data set (guys like LeBron, Steph, Draymond, Duncan, and Ginobili), and even some of those wouldn’t be much less noisy than single-season RAPM (which is super noisy). And it’s actually probably even worse than just looking at the off minutes would lead you to believe, because, even for those players with lots of playoff minutes, the model will be trying to control for the presence of other players that have a very small amount of playoff data. So, basically, the adjustments the model makes compared to on-off will be mostly based on garbage.

.

I agree completely, noisy samples like this or like the great work squared and others did with pre 2000's rapm are nice to have but extremely flawed to actually use in comparision to bigger and complete data sets of other players (or the same player*)

i mean the problem is the larger datasets are all regular season stuff. or at least massively weighted to the regular season stuff, when most people's rankings are heavily playoff based. so using the RAPM from the regular tends and the "rangz" from the playoffs tends to give some players a "have their cake and eat it too" way of ranking people.

[DraymondGold]

it's always interesting that when there is some partial RAPM data released by squared or partial on/off data calculated by somebody or released by squared or somebody else and even if you have incomplete data or small off samples or whatever, as long as those data sets show certain players looking good there is no "cautioning," no "not very useful," "too small amount of data," for the various other caveats that you see not only in this thread, but every thread. After the usual "cautioning" and cook take it with a grain of salt" and "small/tiny off samples," you see long posts talking about why the data isn't useful or not to be taken too seriously even though those very same cautioning or precautions aren't done with other data that also has small and/or incomplete sample sizes. In fact, those are the threats are responded to with "this is amazing work!" "this dara really shows something."

I am 100% sure that had this particular set not had LeBron far away at the top or curry below Draymond or Jokić nowhere near the top which is reflective of his playoffs on/off, etc., these very same cautioning, but of course not be given.

Consistency would be nice.

I’m sorry but this is a bit maddening. It seems fairly clear that you’re at least in part meaning to refer to me, since your post comes after a post that responded to me about this same issue (not to mention that you frequently make snide remarks directed at me, and your post’s other examples of Curry/Draymond and Jokic completely align with this referring to me). But you’ll find that I’m actually really consistent in caveating and cautioning regarding Squared’s data. I can’t even count how many times I’ve done so. Indeed, I actually kind of doubt there’s anyone on these forums who has done so more times than I have. I even just did it in response to Falco’s post, before I saw your post (ironically, I even used some of the exact same language you claim is never used regarding Squared, specifically saying it should be taken with “a grain of salt”)! The idea that “those same cautioning and precautions aren’t done with other data” is just manifestly false if you’re referring to me (which you definitely are). You’ll also find that I’m very consistent about expressing concerns regarding playoff RAPM, not just in this thread. Again, the idea that my concerns in this arena are motivated by the results of any particular measure is nonsense that is completely divorced from my actual post history. It’s just a complete straw man, and I find it maddening because it’s not the first time you’ve made posts like this. You tend to make posts where you are obviously talking about me without explicitly saying you’re talking about me, while making assertions/accusations about my post history that are just objectively untrue. I’d appreciate you not doing so. I can understand that you probably have not have read all my posts and that that could potentially be a good faith cause of your incorrect impressions, but even if that’s the case I’d say that you shouldn’t make broad assertions/accusations about my posts if you haven’t actually done your homework enough to know that what you’re saying is accurate.

The maddening thing is when posters feign neutrality sans agenda when what they post is very often that. I think it's ok to be partisan and out of partisanship we might collect/post new and interesting data like a Cold Era arms race leading to greater tech. Claim it and own it rather than hide behind some sort of objectivity, "i'm being mischaracterized," etc.

This is your first post right after the data release:

Given how tiny playoff off samples are, I think this is mostly extremely noisy to the point of being not very useful, except perhaps to look at the small number of people with tons of playoff games in this data set (guys like LeBron, Steph, Draymond, Duncan, and Ginobili), and even some of those wouldn’t be much less noisy than single-season RAPM (which is super noisy). And it’s actually probably even worse than just looking at the off minutes would lead you to believe, because, even for those players with lots of playoff minutes, the model will be trying to control for the presence of other players that have a very small amount of playoff data. So, basically, the adjustments the model makes compared to on-off will be mostly based on garbage.

There is a straightaway dismissal of the data. There's lots of other data posted that's incomplete, noisy, but for that data, there isn't a straightaway dismissal. I wonder why.

Had this had a different player at the top or if Jokic looked better it Curry over Draymond, etc., I'm confident that the first post would not be a straightaway dismissal. Maybe we'd get



Or from the other poster here feigning the same we'd get



or

instead of a 1,000 word dismissal.

Again, consistency and owning it.

My opening sentence in the last thread that homecourtloss is upset about: “Fantastic stuff, Djoker.”
My opening sentence in this thread that homecourtloss is upset about: “Cool stuff!”
homecourtloss: accuses me of being inconsistent and anti-LeBron biased because of this different opening sentence.



I guess you’re upset that I said cool instead of fantastic? It’s pretty clear that you’re upset about this. No reasonable person would react this way. Ironically you’re the one being inconsistent given how similar the opening sentences are — as any objective person would agree. Genuinely no idea why. You have a history of attacking me and others for some reason.

Given two recent fellow discord members got banned for their sad behavior (which often involved attacks similar to this one), I’d encourage you to take a deep breath, take a step back, and reflect on whether this is really the way you want to engage with fellow realGM members. You certainly are capable of productive and insightful posts. But recently you’ve taken to doing these odd passive aggressive and immature attacks, and defending the actions of two trolls who made alts and tried to organize a pro-LeBron conspiracy.

In the future, I’d encourage you to keep ad hominem attacks out of threads, and focus on the actual basketball content. If not, consider yourself reported and ignored. But I think if would be a shame if a once-productive poster decided this was the way they wanted to engage with others.

[DraymondGold]

OK, based on this I’m immediately deciding that I’ve underrated Kevin Durant.

yeah i've been told KD is a value-less hack who doesn't provide defensive value and can't really run an offense and that he's just a throw in on teams. weird that RAPM doesn't feel the same way. also, i was told james harden's isoball didn't work in the playoffs.

of course, i'm not as big of a believer in RAPM as many other, but for the ones who are, it feels like this should move rankings.

It’s definitely high on Harden! Impressive stuff from him.

Just wondering, how would you rate the playoff performance of Harden compared to the other great players in the play by play era? I’m fine with tiers if that’s easier than exact ranks. Would you say you’re higher or lower on him than this data?

[lessthanjake]

it's always interesting that when there is some partial RAPM data released by squared or partial on/off data calculated by somebody or released by squared or somebody else and even if you have incomplete data or small off samples or whatever, as long as those data sets show certain players looking good there is no "cautioning," no "not very useful," "too small amount of data," for the various other caveats that you see not only in this thread, but every thread. After the usual "cautioning" and cook take it with a grain of salt" and "small/tiny off samples," you see long posts talking about why the data isn't useful or not to be taken too seriously even though those very same cautioning or precautions aren't done with other data that also has small and/or incomplete sample sizes. In fact, those are the threats are responded to with "this is amazing work!" "this dara really shows something."

I am 100% sure that had this particular set not had LeBron far away at the top or curry below Draymond or Jokić nowhere near the top which is reflective of his playoffs on/off, etc., these very same cautioning, but of course not be given.

Consistency would be nice.

I’m sorry but this is a bit maddening. It seems fairly clear that you’re at least in part meaning to refer to me, since your post comes after a post that responded to me about this same issue (not to mention that you frequently make snide remarks directed at me, and your post’s other examples of Curry/Draymond and Jokic completely align with this referring to me). But you’ll find that I’m actually really consistent in caveating and cautioning regarding Squared’s data. I can’t even count how many times I’ve done so. Indeed, I actually kind of doubt there’s anyone on these forums who has done so more times than I have. I even just did it in response to Falco’s post, before I saw your post (ironically, I even used some of the exact same language you claim is never used regarding Squared, specifically saying it should be taken with “a grain of salt”)! The idea that “those same cautioning and precautions aren’t done with other data” is just manifestly false if you’re referring to me (which you definitely are). You’ll also find that I’m very consistent about expressing concerns regarding playoff RAPM, not just in this thread. Again, the idea that my concerns in this arena are motivated by the results of any particular measure is nonsense that is completely divorced from my actual post history. It’s just a complete straw man, and I find it maddening because it’s not the first time you’ve made posts like this. You tend to make posts where you are obviously talking about me without explicitly saying you’re talking about me, while making assertions/accusations about my post history that are just objectively untrue. I’d appreciate you not doing so. I can understand that you probably have not have read all my posts and that that could potentially be a good faith cause of your incorrect impressions, but even if that’s the case I’d say that you shouldn’t make broad assertions/accusations about my posts if you haven’t actually done your homework enough to know that what you’re saying is accurate.

The maddening thing is when posters feign neutrality sans agenda when what they post is very often that. I think it's ok to be partisan and out of partisanship we might collect/post new and interesting data like a Cold Era arms race leading to greater tech. Claim it and own it rather than hide behind some sort of objectivity, "i'm being mischaracterized," etc.

This is your first post right after the data release:

Given how tiny playoff off samples are, I think this is mostly extremely noisy to the point of being not very useful, except perhaps to look at the small number of people with tons of playoff games in this data set (guys like LeBron, Steph, Draymond, Duncan, and Ginobili), and even some of those wouldn’t be much less noisy than single-season RAPM (which is super noisy). And it’s actually probably even worse than just looking at the off minutes would lead you to believe, because, even for those players with lots of playoff minutes, the model will be trying to control for the presence of other players that have a very small amount of playoff data. So, basically, the adjustments the model makes compared to on-off will be mostly based on garbage.

There is a straightaway dismissal of the data. There's lots of other data posted that's incomplete, noisy, but for that data, there isn't a straightaway dismissal. I wonder why.

Had this had a different player at the top or if Jokic looked better it Curry over Draymond, etc., I'm confident that the first post would not be a straightaway dismissal.

This is extremely odd and not at all internally consistent. On one hand, you implicitly point out that I like Steph Curry, by saying I’d have reacted differently if Curry had been above Draymond. But on the other hand, you claim I made “a straightaway dismissal of the data,” when I actually explicitly carved out Steph (and Draymond, LeBron, Duncan, and Ginobili) as a player for whom the data is in a large enough sample that it might actually be useful. It’s just a self-defeating accusation. You said I dismissed it and wouldn’t have dismissed it if it had “Curry over Draymond,” but I actually carved out Curry and Draymond as players for which the data had meaning. Since you correctly identify that Draymond being above Curry is not a result that favors my preferred player, the fact that I carved out those two as having actually meaningful data demonstrates the *complete opposite* of what you are accusing me of.

To the extent you object that, while I carved out almost all the players you actually mention here as being ones with an actually decent sample, I did point out some further issues with the data even for those players, I don’t see you actually disagreeing with the points I made in any substantive way. Just complaining that I made them. And you’ll notice that I went out of my way just above to point out a further issue with Squared’s data. Being rigorous at spotting issues with data is not a bad thing, and you plainly have no basis for an accusation that I only do that with data that doesn’t favor my preferred players.

Anyways, I have to say I find the accusations of selective use of data a bit interesting when you’ve probably thrown hundreds of And1’s at this person (who happened to have posted positively about players you like, and negatively about players you’re not as keen on). One might say the call is coming from inside the house:

Maybe we’d get

That’s a really interesting thing for you to quote here, because you’ll find that Djoker and I went back and forth in that thread regarding the data Djoker presented. After some discussion about how Djoker’s numbers were calculated, I pointed out that Djoker’s spreadsheet listed a lower pace for many series than what was currently listed on Basketball-Reference—which would therefore result in Jordan’s numbers being overstated in Djoker’s data. In other words, when presented with data that was good for a player I like, I delved deeply into the data instead of taking it at face value and then discovered an issue that caused the data to look better for the player I like than it should. So yeah, your attempt to find an example of me taking data I liked completely at face value is actually very good evidence of the contrary. When presented with that data, I actually delved into it and discovered an issue that required the data to be adjusted downwards for the player I like! Of course, after me having pointed this out, I suspect you will still just persist in making snide and completely off-base remarks towards me. But in the off chance that it might actually have any effect on you, I’d once again ask you to stop. The frequent snide drive-by straw-manning you engage in towards me is not appreciated.

[Elpolo_14]

I couldn't have guessed that Jamal Murray would have landed in the TOP 10 and certainly Not above Nikola Jokic

[An Unbiased Fan]

RAPM still being a thing in 2025 is hilarious, almost as much as Manu being rated a better defender than TD lol.

[homecourtloss]

I couldn't have guessed that Jamal Murray would have landed in the TOP 10 and certainly Not above Nikola Jokic

I don't think anybody seriously believes that Murray is more impactful than Jokic, but Jokic's playoff impact profile is weak relative to his regular season impact and also weak relative to the playoffs' impact of some of the other ATGs he gets compared to. This is part of the inconsistency I was talking about: people will put a lot of weight into +/- numbers and declare GOAT this and that but when the same tools (or better tools) point to something else then the data is dismissed outright as noisy, useless, etc. Murray is a playoffs riser who can create his own shot and has been very successful in the playoffs. Just for reference, the Murray ON and Jokic total minutes from 2019 to 2025 is about the same amount of minutes we have for Jordan's playoffs off minutes from 1985 through 1991. Jokic ON and Murray OFF minutes is the same as Jordan's off minutes from 1985 through a series or two in 1997.

This is 2019-2025


This is just 2023:


2020-2025, 2021-2025, 2022-2025, etc., look just as underwhelming for Jokic. Actually, the RAPM doesn't include this last postseason, which would make it worse for Jokic.

2025 playoffs

[Special_Puppy]

I don't think its inconsistent to update your beliefs less when you already have way more information about the player in general. If we had full tracking and play-by-play data for guys before 2000 I think people would give way less weight to the small +/- studies being produced by Squared2020, Djoker, etc even though the underlying quality of those +/- studies didn't really change

[parsnips33]

to me, the biggest drawback is that, just like steph, it's not clear either have the same value without each other. their combined lineups have always been ridiculous, because they are probably the two most synergistic [high level] teammates ever. like usually you would think with player A and player B, their value would be close to:

RAPM A + RAPM B

but for steph and draymond it feels like it's more like

RAPM A + RAPM B + 2 or 3 more points

Probably my favorite thing about the Warriors dynasty, and I completely agree

I've always been curious, are there any historical examples of player combinations that have the same kind of synergistic boost? I wouldn't be surprised if nobody ever came close in terms of degree to Steph/Dray, but surely there are other combos that have a "more than the sum of their parts" effect

[Special_Puppy]

RAPM still being a thing in 2025 is hilarious, almost as much as Manu being rated a better defender than TD lol.

RAPM is a very valuable tool to understanding players, but your confidence in the RAPM should go down the smaller the sample size. The sample size for the playoffs is pretty small for the vast majority of players so you end up getting weird results in a lot of cases.

[falcolombardi]

I couldn't have guessed that Jamal Murray would have landed in the TOP 10 and certainly Not above Nikola Jokic

I don't think anybody seriously believes that Murray is more impactful than Jokic, but Jokic's playoff impact profile is weak relative to his regular season impact and also weak relative to the playoffs' impact of some of the other ATGs he gets compared to. This is part of the inconsistency I was talking about: people will put a lot of weight into +/- numbers and declare GOAT this and that but when the same tools (or better tools) point to something else then the data is dismissed outright as noisy, useless, etc. Murray is a playoffs riser who can create his own shot and has been very successful in the playoffs. Just for reference, the Murray ON and Jokic total minutes from 2019 to 2025 is about the same amount of minutes we have for Jordan's playoffs off minutes from 1985 through 1991. Jokic ON and Murray OFF minutes is the same as Jordan's off minutes from 1985 through a series or two in 1997.

This is 2019-2025


This is just 2023:


2020-2025, 2021-2025, 2022-2025, etc., look just as underwhelming for Jokic. Actually, the RAPM doesn't include this last postseason, which would make it worse for Jokic.

2025 playoffs

Yes but have you considered some people may like the result of the jordan small sample data but not of the jokic small sample data

[lessthanjake]

to me, the biggest drawback is that, just like steph, it's not clear either have the same value without each other. their combined lineups have always been ridiculous, because they are probably the two most synergistic [high level] teammates ever. like usually you would think with player A and player B, their value would be close to:

RAPM A + RAPM B

but for steph and draymond it feels like it's more like

RAPM A + RAPM B + 2 or 3 more points

Probably my favorite thing about the Warriors dynasty, and I completely agree

I've always been curious, are there any historical examples of player combinations that have the same kind of synergistic boost? I wouldn't be surprised if nobody ever came close in terms of degree to Steph/Dray, but surely there are other combos that have a "more than the sum of their parts" effect

I’ve always felt like Duncan and Ginobili had a lot of synergy. Not quite to the same degree as Steph/Draymond, but it’s in the same ballpark IMO. Before Duncan got a bit old, the two of them had some pretty wild net ratings when on the court together. For instance, in the four years from 2004-2007 (i.e. still Duncan’s true prime, but not Ginobili’s rookie year), those two had a +15.79 net rating together in 5491 regular season minutes (and it was +16.87 if we limited things to just the 2005-2007 three-year span). It was +9.06 in the playoffs, and +12.56 if we added in 2003 (where Ginobili was definitely providing a lot of value by the time the playoffs came around). Even if we expand it all the way out to 2003-2010, the two of them had a +13.43 net rating together in the regular season, and +6.03 in the playoffs. Obviously those numbers aren’t quite as good, but that’s an 8-year span!

To me, this sort of thing reflects well on everyone involved. I think people sometimes try to use this sort of synergy to downplay the top-end guy (i.e. Steph and Duncan). But, to me, being able to synergize with another great player such that your teams can scale up to this kind of level with both of you on the court is a really big feather in a player’s cap. Of course, it also reflects well on the lesser stars in these duos as well.

[lessthanjake]

I couldn't have guessed that Jamal Murray would have landed in the TOP 10 and certainly Not above Nikola Jokic

I don't think anybody seriously believes that Murray is more impactful than Jokic, but Jokic's playoff impact profile is weak relative to his regular season impact and also weak relative to the playoffs' impact of some of the other ATGs he gets compared to. This is part of the inconsistency I was talking about: people will put a lot of weight into +/- numbers and declare GOAT this and that but when the same tools (or better tools) point to something else then the data is dismissed outright as noisy, useless, etc. Murray is a playoffs riser who can create his own shot and has been very successful in the playoffs. Just for reference, the Murray ON and Jokic total minutes from 2019 to 2025 is about the same amount of minutes we have for Jordan's playoffs off minutes from 1985 through 1991. Jokic ON and Murray OFF minutes is the same as Jordan's off minutes from 1985 through a series or two in 1997.

This is 2019-2025


This is just 2023:


2020-2025, 2021-2025, 2022-2025, etc., look just as underwhelming for Jokic. Actually, the RAPM doesn't include this last postseason, which would make it worse for Jokic.

2025 playoffs

Yes but have you considered some people may like the result of the jordan small sample data but not of the jokic small sample data

There’s not much of a comparison between those two samples. The data in Engelmann’s playoff RAPM includes 80 playoff games for Jokic. It’s basically one season’s worth of data for him, and playoff data is even more noisy than that for reasons already explained to you in this thread by me and others. The Squared data for Jordan has 391 regular season games for Jordan (and the on-off data I’ve compiled for him has 739 games worth of data). These aren’t similar in terms of being “small sample data.”

There’s also the important point Special_Puppy made, which is that the weight put on data is naturally dependent on the amount of other data that exists. For Jokic, we have tons of regular season data that tells us he is an incredibly impactful player. So a small sample of data is naturally not going to move the needle as much as it would if we had no other data. For players prior to the play-by-play era, Squared’s data is one of the only pieces of impact data we have, so it will of course get more weight, despite its flaws. If we actually had fulsome data for those players, then Squared’s data would cease to be given much of any weight. I don’t think this should be a difficult concept to understand. The weight put on a piece of data when evaluating a player is a function not only of how reliable or flawed that data is but also how much other information there is. In the absence of much other information, the most reasonable approach is to still put weight on flawed data, while acknowledging that we shouldn’t be particularly certain about the conclusions deriving from it. Which is exactly how I and others have approached it. But apparently even that approach is a source of uncontainable fury to certain people.

[homecourtloss]

I couldn't have guessed that Jamal Murray would have landed in the TOP 10 and certainly Not above Nikola Jokic

I don't think anybody seriously believes that Murray is more impactful than Jokic, but Jokic's playoff impact profile is weak relative to his regular season impact and also weak relative to the playoffs' impact of some of the other ATGs he gets compared to. This is part of the inconsistency I was talking about: people will put a lot of weight into +/- numbers and declare GOAT this and that but when the same tools (or better tools) point to something else then the data is dismissed outright as noisy, useless, etc. Murray is a playoffs riser who can create his own shot and has been very successful in the playoffs. Just for reference, the Murray ON and Jokic total minutes from 2019 to 2025 is about the same amount of minutes we have for Jordan's playoffs off minutes from 1985 through 1991. Jokic ON and Murray OFF minutes is the same as Jordan's off minutes from 1985 through a series or two in 1997.

This is 2019-2025


This is just 2023:


2020-2025, 2021-2025, 2022-2025, etc., look just as underwhelming for Jokic. Actually, the RAPM doesn't include this last postseason, which would make it worse for Jokic.

2025 playoffs

Yes but have you considered some people may like the result of the jordan small sample data but not of the jokic small sample data

The sophistry continues.

Contribute something or refrain from posting please. trex

[Doctor MJ]

So I'll echo trex here to start and just emphasize:

We're here to talk ball not talk about each other.

In practice any community is going to spend some time talking about the community and those who are a part of it, but if you find yourself thinking more about the problem you have with a poster's basketball reasoning than about your own basketball reasoning, I'd urge you to take a step back.

Now, at the root of the issues in this particular thread, in my assessment, are some really juicy apparent differences between what RS RAPM seems to say and what this PS RAPM seems to say, and I think we absolutely should talk about those.

I think any approach that doesn't first deal with a) +/- data being noisy b) particularly in small sample of play c) especially when the opponent sample is even smaller than that. Because of this, it's really not clear how much stock we should put in playoff RAPM. I'm super-interested in it and want to talk about the data we see, but I'm also cautious.

Now, in most cases, the RS & PS data largely go along with each other, which makes that PS data more reassuring, but also doesn't necessarily give us much new to talk about. So we naturally focus on what we find to be discrepant between the RS & PS, and I'd say this tend to cause us to zoom in on pairs of teammates.

The three pairs I think we've talked about most on this are Duncan/Ginobili, Curry/Green, & Jokic/Murray. In all 3 cases, there's an MVP level guy (Duncan/Curry/Jokic) with great +/- data in general, but who looks less impressive by PS data than his teammate, and so we're forced to reckon with this.

I want to be clear in all cases that I think I welcome people making the argument that the MVP isn't actually as good as his teammate. I think we should have the intellectual bravery to have those conversations, and this is something I used to do explicitly making threads I tended to call "Talk Me Down" threads. The one that comes to mind right now was one about MVP Derrick Rose vs teammate Luol Deng, and basically asking: Are we sure Deng isn't the better player? And while at the time I tried to make clear that I wasn't looking to convince people to side with Deng, just presenting my own struggle with justifying the consensus belief of Rose > Deng, I have to say... I think I'd side with Deng now that we can look back with hindsight. Rose was the better scorer of course, but Deng was better at most things, and frankly, Rose's scoring doesn't look all that impressive in hindsight.

So I want to question whether Ginobili/Green/Murray are actually better their respective MVP teammates, and more broadly, I'd just like to understand what actually happens on the hardwood regardless of whether that leads to a clear cut choice for the better player.

But I also want to emphasize: The next person here who actually goes on record siding with Ginobili/Green/Murray over their teammates, I think will actually be the first I've seen, and if we can't actually get people to come out on record like that, I do think we should keep in mind that there's a bit of an absurdity to throwing stones.

If we all agree that Duncan/Curry/Jokic deserve to be ranked ahead of their teammates, then what exactly are we disagreeing about basketball-wise?

Now for myself, I have to start with Duncan/Ginobili despite the fact that's probably the least third-rail-ish of the bunch, because I've personally championed Ginobili quite hard in recent years. I'm on record saying that a) I think Ginobili was the better and more valuable offensive player and should have been made the focus of the Spur offense, and that b) I have Ginobili as my POY for the '04-05 season ahead of Duncan.

That's pretty radical stuff to most - feels radical to me as I didn't always believe these things - but I still consider Duncan to be the better overall player when healthy, and most of the time it was Ginobili more so than Duncan with the health issues. The fact that Ginobili looks so good by playoff +/- absolutely helps his cause, but it hasn't been enough for me to side generally with him over Duncan.

The same holds true in the other cases: Forced to choose between these MVPs and their teammates, I end up siding with the MVPs like basically everyone else... but again, not because I think I have to, and not because I haven't considered the alternative.

In the case of Curry/Green, we're literally talking about this magical circumstance where two synergistic paradigm-shifting players emerge on the same team at roughly the same time. It's crazy, and has everything to do with why their team (Warriors) had a run of dominance like few others in history, and which arguably stands out all the more a decade later as we see a different champion every year. In this age of radical roster changes being the norm, it's really hard to keep a dynastic core together.

Further, while Curry isn't necessarily the best offensive player of his generation depending on how you slice "generation", I'll say flat out I consider Green to be the best defensive player of his generation, while also having a higher BBIQ that allows him to be quite valuable on offense even when his shot isn't working.

But aside from the fact that the RS +/- data seems to favor the guy the basketball world sees as the superior player here, there's also the matter that Green really hasn't given us a Deng-like track record of leading a team to overperform expectations when his scoring star is gone for extended time. I personally wonder if Green could have done a lot more in Curry's big injury year than he did and just didn't case to exert himself in a lost cause, but from a proof-of-concept perspective, we've still never seen Green carry a team without mega-scorers, and that certainly makes me cautious.

Meanwhile in the PS +/- data, we're not talking about something utterly overwhelming to Green's advantage. Depending on the slice of data I look at, sometimes Green looks more impressive, sometimes Curry does. It's honestly just not enough to make me flip my opinion that came from the regular season, and I'll say particularly when what we're talking about are two teammates. If these were rivals on competing teams, one guy can actually win while the other loses, but as is, they win & lose together relying on each other the whole time, and given that they've done so much winning, it's hard for me to really knock either of them down.

Jokic/Murray might be the most interesting of the bunch because we seem to see a huge RAPM advantage for Jokic in the regular season, but when Murray has been "on" in the playoffs (2020, 2023), he's absolutely looked like a superstar too, while having that major on-off advantage. Given Murray's inconsistency, it's entirely plausible to argue:

Hot Jamal > Jokic > Cold Jamal

But I'm not comfortable saying this at this time. Obviously I'd like to see more sample like this to ease me passed my caution, but there's a specific note I'll end on:

What do we do with the scenario where a team is built around Player A, opponents focus their strategy around Player A, but Player B appears to be the more impactful player by RAPM?

It's of course possible that Player B is the guy who should have been built around, but I would argue that oftentimes what's going on is a kind of "stealth impact" that comes from the opponent not optimizing against Player B, and thus Player B's stealth impact can be considered "fragile" to some degree compared to a guy who is doing his thing against the teeth of the defense's attention.

The classic extreme example of this can be found in the RS data, and that's Nick Collison in OKC. The +/- data sees him as impactful as his star teammates, yet he played limited minutes for reasons that didn't seem to be about poor conditioning. For a while it was reasonable to think the data was noise, but it lasted long enough I don't feel comfortable chalking it up to that.

Rather, I see Collison's +/- impact as being more about him being put on the court in contexts where he could max out stealth impact by virtue of a) a great fit in the Thunder lineup, and b) opponent focus on his teammates. Which is another way of saying that if you play Collison huge minutes with the expectation he can keep that impact up, I think it backfires hard.

In the case of Murray, we're talking about a star-level player who does get opponent attention like a #2 guy, but aside from that #2 is not #1, there's also the matter that the way the Nuggets play around Jokic is so unusual by modern NBA standards, and so the way playoff opponents look to battle against Jokic's team strategically is just not the same way you'd play if you were looking to combat an offense led by a scoring guard with non-elite passing.

This then to say that I don't really think there's reason to believe that Murray on a random team as their alpha is going to be some insane playoff performer every few years that makes his team top contenders more effectively than Jokic.

But if others disagree, I'd love to hear how people are thinking about the basketball here.