’94 Hakeem Olajuwon vs ’23 Nikola Jokic

[migya]

Recently watched Olajuwon in 94 playoffs and his offense was elite, scoring and passing. He is not behind any other center.

If you just watched 1994 Hakeem and concluded that his passing is not behind Jokic, then I suggest trying to watch Jokic now...

I assume he meant in general

Like if someone said

“I saw highlights of peak bron, his defense, scoring, passing were so good man. No ones better than him”

I wouldn’t conclude that person thinks bron > Hakeem on defense know what I mean

I meant what I wrote. Olajuwon's offense was elite and against the best defending bigs. He had less shooting around him and got the ball to them. He scored on insane moves and shot very well. Defense is not a discussion here.

[MyUniBroDavis]

If you just watched 1994 Hakeem and concluded that his passing is not behind Jokic, then I suggest trying to watch Jokic now...

I assume he meant in general

Like if someone said

“I saw highlights of peak bron, his defense, scoring, passing were so good man. No ones better than him”

I wouldn’t conclude that person thinks bron > Hakeem on defense know what I mean

I meant what I wrote. Olajuwon's offense was elite and against the best defending bigs. He had less shooting around him and Dot the ball to them. He scored on insane moves and shot very well. Defense is not a discussion here.

Hakeem was a great offensive player, and overall I have hakeem here pretty easily, but I’m sure ur not trying to say he’s a comparable passer to Jokic right lol

[migya]

I assume he meant in general

Like if someone said

“I saw highlights of peak bron, his defense, scoring, passing were so good man. No ones better than him”

I wouldn’t conclude that person thinks bron > Hakeem on defense know what I mean

I meant what I wrote. Olajuwon's offense was elite and against the best defending bigs. He had less shooting around him and Dot the ball to them. He scored on insane moves and shot very well. Defense is not a discussion here.

Hakeem was a great offensive player but I’m sure ur not trying to say he’s a comparable passer to Jokic right lol

I wrote offense, so overall offensively.

[70sFan]

If you just watched 1994 Hakeem and concluded that his passing is not behind Jokic, then I suggest trying to watch Jokic now...

I assume he meant in general

Like if someone said

“I saw highlights of peak bron, his defense, scoring, passing were so good man. No ones better than him”

I wouldn’t conclude that person thinks bron > Hakeem on defense know what I mean

I meant what I wrote. Olajuwon's offense was elite and against the best defending bigs. He had less shooting around him and Dot the ball to them. He scored on insane moves and shot very well. Defense is not a discussion here.

There are only two things Hakeem did potentially better than Jokic on offense:

- lob finishing (which is not really that important),
- isolation post scoring (which is questionable really, because Jokic himself is an all-time great post scorer).

That's not enough to outweigh the massive advantage Jokic has in passing, shooting, finishing and off-ball play.

[lessthanjake]

Is there a statistical basis for the higher views on Hakeem that some people have? I’m sure there’s metrics I am not aware of and there’s metrics that are behind paywalls, so this is a genuine question. People seem to have very strong views about him, but there’s not actually a lot of advanced metrics that go far enough back to capture his impact one way or the other, and things I am aware that we do have such as WOWYR are not actually very high on him. Based on the information I’m aware of, the case for thinking really highly of Hakeem (and I’d classify thinking 2023 Jokic is nowhere near him to be thinking quite highly of him, given how great Jokic was) can’t really be based on advanced metrics. Which is fine—there’s other criteria one can use and advanced metrics are flawed and have blind spots, so there’s nothing inherently wrong with relying more on one’s intuitive sense from watching a player. But it’s just a bit surprising to me how high most people on this particular sub-forum seem to be on him, when I see his case as being less metric-based than most. It makes me feel like I must be missing some source of data.

[AEnigma]

Well, RealGM crashed my post, so this will be kept briefer.

To start, WOWYR is a messy metric that can swing wildly with slight input adjustments or changes in approach, but all the same, Hakeem tends to post values similar to or even better than players like Duncan, Wilt, and Bird — and he has among the absolute best WOWY indicators.


You can go read the Backpicks / Thinking Basketball career profile on him, or watch the video making the case for him as a top four peak, and they can touch on how a lot of those typical metrics — even ones which try to incorporate on/off like AuPM does — have a tendency to undersell defensive contributions, and how Hakeem’s defensive results without all-time defensive talent or coaching speak to a level of impact on that end which goes beyond the raw rDrtgs of his teams.

However, you asked specifically about “impact”, and there I think it is essential to look at Hakeem as a player whose teams consistently (albeit not universally) elevate in the postseason, typically scaling right in line with his own increase in box productivity. If the 1993-95 Rockets seem like “underwhelming” regular season efforts relative to what you see from the lift of players like Garnett, I think it looks pretty damn good when you add an extra five to ten wins in expected postseason performance.

You can see an echo of that elevation (or as Ben would term in, “inelasticity”) in the regular season too, where Hakeem’s teams consistently (again, not universally) outperform their win expectations based off point differential. Lebron is the easy modern analogy there: basically, if a game is close, some players are more likely to “steal” wins, even if that value is not something you can realistically capture in WOWY or its derivatives.

Prime Hakeem is maybe only like a top ten to fifteen regular season player… but then he elevates in the postseason while almost everyone else falls off or holds firm. Oh, and he has strong fifteen-year longevity, which is a fair bit more than guys like Curry, Bird, Jordan, and Magic. I do not find the case that oblique, and while I recognise he is in a sense disadvantaged by not having playoff on/off for people to cite, I think it should be easy to see the potential impact elevation in play from someone routinely overachieving in the postseason without much surrounding talent to bleed away the assigned value of that overachievement.

[lessthanjake]

Well, RealGM crashed my post, so this will be kept briefer.

To start, WOWYR is a messy metric that can swing wildly with slight input adjustments or changes in approach, but all the same, Hakeem tends to post values similar to or even better than players like Duncan, Wilt, and Bird — and he has among the absolute best WOWY indicators.


You can go read the Backpicks / Thinking Basketball career profile on him, or watch the video making the case for him as a top four peak, and they can touch on how a lot of those typical metrics — even ones which try to incorporate on/off like AuPM does — have a tendency to undersell defensive contributions, and how Hakeem’s defensive results without all-time defensive talent or coaching speak to a level of impact on that end which goes beyond the raw rDrtgs of his teams.

However, you asked specifically about “impact”, and there I think it is essential to look at Hakeem as a player whose teams consistently (albeit not universally) elevate in the postseason, typically scaling right in line with his own increase in box productivity. If the 1993-95 Rockets seem like “underwhelming” regular season efforts relative to what you see from the lift of players like Garnett, I think it looks pretty damn good when you add an extra five to ten wins in expected postseason performance.

You can see an echo of that elevation (or as Ben would term in, “inelasticity”) in the regular season too, where Hakeem’s teams consistently (again, not universally) outperform their win expectations based off point differential. Lebron is the easy modern analogy there: basically, if a game is close, some players are more likely to “steal” wins, even if that value is not something you can realistically capture in WOWY or its derivatives.

Prime Hakeem is maybe only like a top ten to fifteen regular season player… but then he elevates in the postseason while almost everyone else falls off or holds firm. Oh, and he has strong fifteen-year longevity, which is a fair bit more than guys like Curry, Bird, Jordan, and Magic. I do not find the case that oblique, and while I recognise he is in a sense disadvantaged by not having playoff on/off for people to cite, I think it should be easy to see the potential impact elevation in play from someone routinely overachieving in the postseason without much surrounding talent to bleed away the assigned value of that overachievement.

Okay, so I was aware of all of that already, so I guess maybe there isn’t anything else. And I think the bottom line is that the case for Hakeem is simply not really an advanced-metric-based case. It’s a case that relies on saying things like “Look who he won titles with” and “Yeah but a lot of advanced metrics may not do a great job measuring defensive impact, and I weigh Hakeem’s great defense quite highly.” Those are much squishier evaluations than an advanced-metric-based case. And that’s fine! Evaluating basketball greatness really isn’t just about hard data—which doesn’t encompass everything important, and is inherently prone to all the same kinds errors that any data analysis is (just because someone ran a regression doesn’t mean the data is good or that the methodology is sound). But, given the fact that it’s not really a data-based case, I’m a little surprised at the fact that many people on this particular sub-forum seem to have such certainty at a really high evaluation of Hakeem. And, I do tend to wonder whether there’s a bit of groupthink going on, with essentially every poster here having consumed Ben Taylor’s content and all perhaps simultaneously being influenced by his conclusions. For instance, Ben Taylor is higher on both Hakeem and Garnett than the general norm, albeit actually for different reasons and based on different evidence and thought processes, and I’ve noticed in the personal ranking thread and elsewhere that those two players happened to be consistently ranked surprisingly high by people. And it seems like the reasoning at least for Hakeem (albeit not for Garnett) is not a data-based case, which is not actually typical for this sub-forum. So it does make me wonder about groupthink (though of course Ben Taylor is a great analyst and so I’m certainly not saying his conclusions are wrong or not well-considered). This is not specific to Hakeem to be honest—more just a general point/caution about groupthink when a set of people are all influenced by similar content/opinions.

[70sFan]

What "advanced stats" are we talking about? Boxscore composites available on Basketball-Reference?

Also, Ben Taylor is criticized by all kind of posters here (often undeservedly in my opinion). I have no idea how you can come up with the conclusion that people are heavily influenced by his rankings, when every time Ben posts a new ranking the PC Board is full of unpleasant comments towards it.

[lessthanjake]

What "advanced stats" are we talking about? Boxscore composites available on Basketball-Reference?

Also, Ben Taylor is criticized by all kind of posters here (often undeservedly in my opinion). I have no idea how you can come up with the conclusion that people are heavily influenced by his rankings, when every time Ben posts a new ranking the PC Board is full of unpleasant comments towards it.

To some degree I’m not talking about any specific advanced stat as much as the lack of advanced stats—particularly ones that require tracking data (so, more like impact metrics than box-score-based ones, to answer your question). There’s not a lot of advanced stats with data that goes back far enough to really measure Hakeem’s impact (at least that I’m aware of—that’s why I asked if there was more!), and the ones that do are definitely not uniformly high on him to the level that peoples’ conclusions here seem to be on average. Nor would I say that more box-score-based metrics are all *that* high on him either. So I’d say that the case is clearly not based on advanced metrics. And, as I said, that’s fine, since there’s certainly other valid criteria upon which to base one’s opinion.

As for Ben Taylor specifically, obviously there’s going to be criticisms about anything. And it’s an internet forum—people are *way* more likely to actually post about something if they disagree with it, rather than if they agree or were persuaded by it. I’m also not suggesting *everyone* agrees with him or is influenced by him. And I’m not just talking about Ben Taylor in any event—just a good example that came to mind, but it’s a broader point. In any event, he did rankings that I think are more high on Hakeem and Garnett (to name two examples) than I think the general norm is, and I’ve noticed that a lot of peoples’ personal lists here—a place where I think we can expect most people have consumed Ben Taylor’s content—are also higher on Hakeem and Garnett than the general norm is. And the reasons for being high on the two players aren’t really the same (for Garnett the case centers a lot on off-the-charts impact metrics, while for Hakeem it’s a necessarily more squishy case), so I don’t know that it can be explained as people simply using the same methodology/thought-process to get to those conclusions (though admittedly there’s a general commonality of being two-way impact players). It *feels* to me like it may be a bit of groupthink. I’m not suggesting it’s some uniform phenomenon that is true of every single person here, but it feels to me like it’s had an effect overall—and that shouldn’t really be a surprise when you have a group of people who have mostly consumed very similar content.

[AEnigma]

“This forum dedicated to people thoroughly researching and analysing players at a level (ostensibly) beyond counting up rings or win shares / VORP seems to have a general sense of agreement that this player is really good. Sounds awfully like a case of groupthink.”

But even then, it does not seem you are actually interested in the numbers. To me, when you ask for a metric, and I provide a regular season metric for a high longevity playoff elevator suggesting he had a top ten prime, that would probably indicate a top ten placement. And when other metrics make him look comparable to guys you have a full tier ahead of him, while again having a playoff and/or longevity advantage, someone legitimately curious would probably give that some thought.

Not you though! Guess it really does all come back to ring-counting.

[henshao]

Man you should see how personal some of the disagreements get here, if anything Hakeem being highly rated by such a diverse group should lead you the opposite conclusion of groupthink (there is the chance the guy was really good) I am reminded of a situation involving poker players...

In 1970 poker made headlines all over the United States. Binion, who was temporarily out of jail, had called upon all the best players in the country to come and gather in Las Vegas.

He called the event the “World Series of Poker. Moss had cemented a reputation of being a top-class player and his name was one of the best-known in poker.

Moss won the first World Series, although “winning” probably isn’t the correct term for what happened. The game wasn’t played in a freeze-out format but as a cash game. And all seven(!) participants were asked afterwards to vote for the best player.

With the typical modesty that permeates so many poker players, everybody voted for himself. To determine a winner somehow, they agreed on having a second poll.

This time they voted for who they thought was the second-best player. Moss received the most votes.

https://www.pokerlistings.com/wsop-stories-mr-a-t-the-best-2nd-best-player-in-history

[lessthanjake]

“This forums dedicated to people thoroughly researching and analysing players at a level (ostensibly) beyond counting up rings or win shares / VORP seems to have a general sense of agreement that this player is really good. Sounds awfully like a case of groupthink.”

But even then, it does not seem you are actually interested in the numbers. To me, when you ask for a metric, and I provide a regular season metric for a high longevity playoff elevator suggesting he had a top ten prime, that would probably indicate a top ten placement. And when other metrics make him look comparable to guys you have a full tier ahead of him, while again having a playoff and/or longevity advantage, someone legitimately curious would probably give that some thought.

Not you though! Guess it really does all come back to ring-counting.

You didn’t provide any numbers except a prime WOWY score that has Hakeem 10th. But, as I imagine you are aware, prime WOWYR has him 49th, and 5-year AuPM has him 61st, while box-score-based metrics are not *crazy* high either. Looking at the overall picture, it is not really of a player who has a metric-based case for being ranked super high, unless there’s metrics I don’t know about, but I asked if there were and you did not provide me with anything I was not already aware of. And not having a great metrics-based case is fine! Metrics are flawed and don’t show the whole picture, and metrics are even more limited for a player whose career was before tracking data since there’s fewer metrics to look at.

Anyways, not sure where you get off saying that “it really does all come back to ring-counting” for me, since I didn’t say anything suggestive of that, and you can look at my personal ranking I put in the thread on the personal list thread and it’s plainly not merely a ring-counting list. I think you’re just wanting to be a bit personal here, which is disappointing.

[AEnigma]

“This forums dedicated to people thoroughly researching and analysing players at a level (ostensibly) beyond counting up rings or win shares / VORP seems to have a general sense of agreement that this player is really good. Sounds awfully like a case of groupthink.”

But even then, it does not seem you are actually interested in the numbers. To me, when you ask for a metric, and I provide a regular season metric for a high longevity playoff elevator suggesting he had a top ten prime, that would probably indicate a top ten placement. And when other metrics make him look comparable to guys you have a full tier ahead of him, while again having a playoff and/or longevity advantage, someone legitimately curious would probably give that some thought.

Not you though! Guess it really does all come back to ring-counting.

You didn’t provide any numbers except a prime WOWY score that has Hakeem 10th.

I also suggested you look at playoff elevation numbers where he is uniquely high, but evidently you would prefer some pre-existing ranking to give it much validity.

But, as I imagine you are aware, prime WOWYR has him 49th,

… ahead of Tim Duncan, Kevin Durant, Larry Bird, Moses Malone…

and 5-year AuPM has him 61st,

… which if you had actually bothered to read the article was contextualised as a limited representation of defensive bigs like Hakeem in much the same way 2006-10 Duncan is also undersold by that metric. Nevertheless, even if you had not bothered to read, you probably should be able to pick up on the issue of not having any kind of plus/minus data before 1994 (and generally none in the playoffs before 1997).

while box-score-based metrics are not *crazy* high either.

Yes, stunning how box metrics fall short of properly assessing top tier defenders.

Even with that, though, he is fourth among retired players in playoff BPM (and one of the three ahead is McGrady).

And among big men?



What a disappointment.

Looking at the overall picture, it is not really of a player who has a metric-based case for being ranked super high, unless there’s metrics I don’t know about, but I asked if there were and you did not provide me with anything I was not already aware of. And not having a great metrics-based case is fine! Metrics are flawed and don’t show the whole picture, and metrics are even more limited for a player whose career was before tracking data since there’s fewer metrics to look at.

And so in response to there not being some clear and easily accessible all-in one metric that everyone is using to make the basis of a case — or at least, not to whatever standard you seem to arbitrarily want — you accuse everyone of “groupthink”.

Anyways, not sure where you get off saying that “it really does all come back to ring-counting” for me, since I didn’t say anything suggestive of that, and you can look at my personal ranking I put in the thread on the personal list thread and it’s plainly not merely a ring-counting list. I think you’re just wanting to be a bit personal here, which is disappointing.

It is not as much of a mystery as you want to portray it why you think there are eleven players clearly better than Hakeem. Tough to see mere coincidence in you throwing every top player with three rings ahead of him and trying to isolate his Finals runs as merely lucky outliers in a career of early exits.

[lessthanjake]

“This forums dedicated to people thoroughly researching and analysing players at a level (ostensibly) beyond counting up rings or win shares / VORP seems to have a general sense of agreement that this player is really good. Sounds awfully like a case of groupthink.”

But even then, it does not seem you are actually interested in the numbers. To me, when you ask for a metric, and I provide a regular season metric for a high longevity playoff elevator suggesting he had a top ten prime, that would probably indicate a top ten placement. And when other metrics make him look comparable to guys you have a full tier ahead of him, while again having a playoff and/or longevity advantage, someone legitimately curious would probably give that some thought.

Not you though! Guess it really does all come back to ring-counting.

You didn’t provide any numbers except a prime WOWY score that has Hakeem 10th.

I also suggested you look at playoff elevation numbers where he is uniquely high, but evidently you would prefer some pre-existing ranking to give it much validity.

Playoff elevation isn’t an advanced metric. It’s just an observation that his box-score numbers have gone up in the playoffs. Which is a very good thing, and should weigh into any ranking of him of course (maybe a lot!), but I was talking about whether he had an advanced-metric-based case for being ranked so high, and this isn’t that. So I’m not sure why you are getting so aggressive/sarcastic towards me about that in the context of this specific discussion. Please try to be civil and avoid launching barrages of sarcastic remarks over a basketball discussion.

But, as I imagine you are aware, prime WOWYR has him 49th,

… ahead of Tim Duncan, Kevin Durant, Larry Bird, Moses Malone…

Yes, and obviously behind lots of other people. If he’s ranked 49th in an advanced metric and is still ahead of those guys, then maybe those guys don’t have a great advanced-metric case either! The point is that no one is going to look at someone being ranked 49th in an advanced ranking and say that that is why they ranked the player way higher than 49th. That’s manifestly obvious.

and 5-year AuPM has him 61st,

… which if you had actually bothered to read the article was contextualised as a limited representation of defensive bigs like Hakeem in much the same way 2006-10 Duncan is undersold hy that metric. Even if you had not bothered to read, though, if you had knowledge of those metrics, you should probably be able to pick up on the issue of not having any kind of plus/minus data before 1994 (and generally none in the playoff before 1997).

All you’re saying is that it was contextualized with reasons to not put much weight on the advanced metrics in this case. That’s not an advanced-metric-based case! That’s a rationale for having the ranking depart from the advanced metrics! So it’s very hard to see what your point is. Saying “This player is really good in a way that is not adequately picked up by the advanced metrics so I rank him higher than the advanced metrics do” is truly the opposite of an advanced-metrics-based case!

while box-score-based metrics are not *crazy* high either.

Yes, stunning how box metrics fall short of assessing defenders.

Again, this is my point! Saying that advanced metrics don’t assess defense very well and Hakeem was a great defender so you rank him higher than the advanced stats is the opposite of an advanced-stats-based argument. It’s an argument explicitly centered on your evaluation of the player’s defense and what you think its value/effect was, rather than on empirical data. And that’s fine! But I don’t know why you’d then turn around and make sarcastic remarks at me for saying the case for Hakeem isn’t a metrics-based case.

Looking at the overall picture, it is not really of a player who has a metric-based case for being ranked super high, unless there’s metrics I don’t know about, but I asked if there were and you did not provide me with anything I was not already aware of. And not having a great metrics-based case is fine! Metrics are flawed and don’t show the whole picture, and metrics are even more limited for a player whose career was before tracking data since there’s fewer metrics to look at.

And so in response to there not being some clear and easily accessible all-in one metric that everyone is using to make a basis of a case — or at least, not to whatever standard you seem to arbitrarily want — you accuse everyone of “groupthink”.

Look, this subforum involves a lot of metrics-based analysis. The case for Hakeem is largely not a metrics-based one, but he seems highly ranked on this site compared to general consensus on him. And the reasons people (including you!) give for ranking him so highly tend to track very very closely with Ben Taylor’s rationale for doing so (indeed sometimes explicitly referring to it, as you did), and Ben Taylor’s analysis is something I’m sure almost everyone here has seen. It’s not crazy to wonder if there’s some groupthink at play, especially when a similar fact pattern exists beyond just Hakeem. Or maybe Ben Taylor’s analysis is just right and people independently came to the same conclusions! I don’t really care much either way, and I don’t want to go in a rabbit hole discussing this. No one is going to think their views are in part a product of groupthink, and it’s not like I have reason to have a *really high* degree of confidence that it is, so it’d just result in a merry-go-round of unfriendly discussion. I shouldn’t actually have brought it up.

Anyways, not sure where you get off saying that “it really does all come back to ring-counting” for me, since I didn’t say anything suggestive of that, and you can look at my personal ranking I put in the thread on the personal list thread and it’s plainly not merely a ring-counting list. I think you’re just wanting to be a bit personal here, which is disappointing.

It is not as much of a mystery as you want to portray it why you think there are eleven players clearly better than Hakeem. Tough to see mere coincidence in you throwing every top player with three rings ahead of him and trying to isolate his Finals runs as merely lucky outliers in a career of early exits.

No, it’s not a mystery. Titles matter in a ranking of “greatness.” They’re not dispositive, otherwise, for instance, I wouldn’t have Steve Nash above Isiah Thomas. But, to me, a significant element of “greatness” in a sport is about raw achievement in that sport (i.e. winning). Thus, someone can be “greater” than a player that is a better player. You might disagree and put little to no weight on team success in an assessment of “greatness.” And that’s fine too! Just another example of how measuring greatness in basketball is a subjective exercise.

[AEnigma]

You didn’t provide any numbers except a prime WOWY score that has Hakeem 10th.

I also suggested you look at playoff elevation numbers where he is uniquely high, but evidently you would prefer some pre-existing ranking to give it much validity.

Playoff elevation isn’t an advanced metric. It’s just an observation that his box-score numbers have gone up in the playoffs. Which is a very good thing, and should weigh into any ranking of him of course (maybe a lot!), but I was talking about whether he had an advanced-metric-based case for being ranked so high, and this isn’t that. So I’m not sure why you are getting so aggressive/sarcastic towards me about that in the context of this specific discussion. Please try to be civil and avoid launching barrages of sarcastic remarks over a basketball discussion.

You said metrics generally, I said numbers; this is pedantry, and “advanced” is a buzzword. If I say all his composites — which you do treat as “metrics” — improve to a degree not true of any other top player, that refers to something measurable.

But, as I imagine you are aware, prime WOWYR has him 49th,

… ahead of Tim Duncan, Kevin Durant, Larry Bird, Moses Malone…

Yes, and obviously behind lots of other people. If he’s ranked 49th in an advanced metric and is still ahead of those guys, then maybe those guys don’t have a great advanced-metric case either! The point is that no one is going to look at someone being ranked 49th in an advanced ranking and say that that is why they ranked the player way higher than 49th. That’s manifestly obvious.

What is manifestly obvious is that you are the type of person who sees 2009 Odom leading the league in RAPM and concludes that means RAPM says he was the #1 player that year. What is manifestly obvious is that you are more interested in doing a bit than actually engaging with what any of this data might suggest.

and 5-year AuPM has him 61st,

… which if you had actually bothered to read the article was contextualised as a limited representation of defensive bigs like Hakeem in much the same way 2006-10 Duncan is undersold hy that metric. Even if you had not bothered to read, though, if you had knowledge of those metrics, you should probably be able to pick up on the issue of not having any kind of plus/minus data before 1994 (and generally none in the playoff before 1997).

All you’re saying is that it was contextualized with reasons to not put much weight on the advanced metrics in this case. That’s not an advanced-metric-based case! That’s a rationale for having the ranking depart from the advanced metrics! So it’s very hard to see what your point is. Saying “This player is really good in a way that is not adequately picked up by the advanced metrics so I rank him higher than the advanced metrics do” is truly the opposite of an advanced-metrics-based case!

No, it is saying if you are going to use these metrics, know how to use them properly.

Apparently that is asking too much.

while box-score-based metrics are not *crazy* high either.

Yes, stunning how box metrics fall short of assessing defenders.

Again, this is my point! Saying that advanced metrics don’t assess defense very well and Hakeem was a great defender so you rank him higher than the advanced stats is the opposite of an advanced-stats-based argument. It’s an argument explicitly centered on your evaluation of the player’s defense and what you think its value/effect was, rather than on empirical data. And that’s fine! But I don’t know why you’d then turn around and make sarcastic remarks at me for saying the case for Hakeem isn’t a metrics-based case.

Because again it is a consequence of you not bothering to engage with the meaning of the metrics and instead just looking at them as a cold ranking. Roles matter, and even the people who put those metrics together are very clear about that. The problem is not the measurement itself, but how people like you reflexively use them in the shallowest possible way.

Looking at the overall picture, it is not really of a player who has a metric-based case for being ranked super high, unless there’s metrics I don’t know about, but I asked if there were and you did not provide me with anything I was not already aware of. And not having a great metrics-based case is fine! Metrics are flawed and don’t show the whole picture, and metrics are even more limited for a player whose career was before tracking data since there’s fewer metrics to look at.

And so in response to there not being some clear and easily accessible all-in one metric that everyone is using to make a basis of a case — or at least, not to whatever standard you seem to arbitrarily want — you accuse everyone of “groupthink”.

Look, this subforum involves a lot of metrics-based analysis. The case for Hakeem is largely not a metrics-based one, but he seems highly ranked on this site compared to general consensus on him. And the reasons people (including you!) give for ranking him so highly tend to track very very closely with Ben Taylor’s rationale for doing so (indeed sometimes explicitly referring to it, as you did), and Ben Taylor’s analysis is something I’m sure almost everyone here has seen. It’s not crazy to wonder if there’s some groupthink at play, especially when a similar fact pattern exists beyond just Hakeem.

Or maybe you are baselessly assuming that the majority of people here are coldly going off some collage of various all-in-ones rather than trying to understand and assess the sport.

Or maybe Ben Taylor’s analysis is just right and people independently came to the same conclusions! I don’t really care much either way, and I don’t want to go in a rabbit hole discussing this. No one is going to think their views are in part a product of groupthink, and it’s not like I have reason to have a *really high* degree of confidence that it is, so it’d just result in a merry-go-round of unfriendly discussion. I shouldn’t actually have brought it up.

Yeah, people were able to watch and assess Hakeem’s career before Ben enlightened us all. Same with Garnett, and same with Reggie, and same with Nash. Those takes were shared here long before he publicised them. To the extent that publication is a net benefit to discussion, it is that more people can see the reasoning and potentially revisit those players with a new lens (regardless of whether they ultimately agree). But he was never on an island.

Anyways, not sure where you get off saying that “it really does all come back to ring-counting” for me, since I didn’t say anything suggestive of that, and you can look at my personal ranking I put in the thread on the personal list thread and it’s plainly not merely a ring-counting list. I think you’re just wanting to be a bit personal here, which is disappointing.

It is not as much of a mystery as you want to portray it why you think there are eleven players clearly better than Hakeem. Tough to see mere coincidence in you throwing every top player with three rings ahead of him and trying to isolate his Finals runs as merely lucky outliers in a career of early exits.

No, it’s not a mystery. Titles matter in a ranking of “greatness.” They’re not dispositive, otherwise, for instance, I wouldn’t have Steve Nash above Isiah Thomas. But, to me, a significant element of “greatness” in a sport is about raw achievement in that sport (i.e. winning). Thus, someone can be “greater” than a player that is a better player. You might disagree and put little to no weight on team success in an assessment of “greatness.” And that’s fine too! Just another example of how measuring greatness in basketball is a subjective exercise.

Yes but do you see why it does not work to object when I say this ultimately comes back to you not being able to get over Hakeem having only two titles.

[lessthanjake]

I also suggested you look at playoff elevation numbers where he is uniquely high, but evidently you would prefer some pre-existing ranking to give it much validity.

Playoff elevation isn’t an advanced metric. It’s just an observation that his box-score numbers have gone up in the playoffs. Which is a very good thing, and should weigh into any ranking of him of course (maybe a lot!), but I was talking about whether he had an advanced-metric-based case for being ranked so high, and this isn’t that. So I’m not sure why you are getting so aggressive/sarcastic towards me about that in the context of this specific discussion. Please try to be civil and avoid launching barrages of sarcastic remarks over a basketball discussion.

You said metrics generally, I said numbers; this is pedantry, and “advanced” is a buzzword. If I say all his composites — which you do treat as “metrics” — improve to a degree not true of any other top player, that refers to something measurable.

… ahead of Tim Duncan, Kevin Durant, Larry Bird, Moses Malone…

Yes, and obviously behind lots of other people. If he’s ranked 49th in an advanced metric and is still ahead of those guys, then maybe those guys don’t have a great advanced-metric case either! The point is that no one is going to look at someone being ranked 49th in an advanced ranking and say that that is why they ranked the player way higher than 49th. That’s manifestly obvious.

What is manifestly obvious is that you are the type of person who sees 2009 Odom leading the league in RAPM and concluding that means RAPM says he was the #1 player that year. What is manifestly obvious is that you are more interested in doing a bit than actually trying to engaging with what any of this data might suggest.

… which if you had actually bothered to read the article was contextualised as a limited representation of defensive bigs like Hakeem in much the same way 2006-10 Duncan is undersold hy that metric. Even if you had not bothered to read, though, if you had knowledge of those metrics, you should probably be able to pick up on the issue of not having any kind of plus/minus data before 1994 (and generally none in the playoff before 1997).

All you’re saying is that it was contextualized with reasons to not put much weight on the advanced metrics in this case. That’s not an advanced-metric-based case! That’s a rationale for having the ranking depart from the advanced metrics! So it’s very hard to see what your point is. Saying “This player is really good in a way that is not adequately picked up by the advanced metrics so I rank him higher than the advanced metrics do” is truly the opposite of an advanced-metrics-based case!

No, it is saying if you are going to use this metrics, know how to use then properly.

Apparently that is asking too much.

Yes, stunning how box metrics fall short of assessing defenders.

Again, this is my point! Saying that advanced metrics don’t assess defense very well and Hakeem was a great defender so you rank him higher than the advanced stats is the opposite of an advanced-stats-based argument. It’s an argument explicitly centered on your evaluation of the player’s defense and what you think its value/effect was, rather than on empirical data. And that’s fine! But I don’t know why you’d then turn around and make sarcastic remarks at me for saying the case for Hakeem isn’t a metrics-based case.

Because again it is a consequence of you not bothering to engage with the meaning of the metrics and instead just looking at them as a cold ranking. Roles matter, and even the people who put those metrics together are very clear about that. The problem is not the measurement itself, but how people like you reflexively use them in the shallowest possible way.

And so in response to there not being some clear and easily accessible all-in one metric that everyone is using to make a basis of a case — or at least, not to whatever standard you seem to arbitrarily want — you accuse everyone of “groupthink”.

Look, this subforum involves a lot of metrics-based analysis. The case for Hakeem is largely not a metrics-based one, but he seems highly ranked on this site compared to general consensus on him. And the reasons people (including you!) give for ranking him so highly tend to track very very closely with Ben Taylor’s rationale for doing so (indeed sometimes explicitly referring to it, as you did), and Ben Taylor’s analysis is something I’m sure almost everyone here has seen. It’s not crazy to wonder if there’s some groupthink at play, especially when a similar fact pattern exists beyond just Hakeem.

You maybe you just baselessly assuming that the majority of people here are coldly going off some collage of various all-in-ones rather than trying to understand and assess the sport.

Or maybe Ben Taylor’s analysis is just right and people independently came to the same conclusions! I don’t really care much either way, and I don’t want to go in a rabbit hole discussing this. No one is going to think their views are in part a product of groupthink, and it’s not like I have reason to have a *really high* degree of confidence that it is, so it’d just result in a merry-go-round of unfriendly discussion. I shouldn’t actually have brought it up.

Yeah, people were able to watch and assess Hakeem’s career before Ben enlightened us all. Same with Garnett, and same with Reggie, and same with Nash. Those takes were shared here long before he publicised them. To the extent that publication is a net benefit to discussion, it is that more people can see the reasoning and potentially revisit those players with a new lens (regardless of whether they ultimately agree). But he was never on an island.

It is not as much of a mystery as you want to portray it why you think there are eleven players clearly better than Hakeem. Tough to see mere coincidence in you throwing every top player with three rings ahead of him and trying to isolate his Finals runs as merely lucky outliers in a career of early exits.

No, it’s not a mystery. Titles matter in a ranking of “greatness.” They’re not dispositive, otherwise, for instance, I wouldn’t have Steve Nash above Isiah Thomas. But, to me, a significant element of “greatness” in a sport is about raw achievement in that sport (i.e. winning). Thus, someone can be “greater” than a player that is a better player. You might disagree and put little to no weight on team success in an assessment of “greatness.” And that’s fine too! Just another example of how measuring greatness in basketball is a subjective exercise.

Yes but do you see why it does not work to object when I say this ultimately comes back to you not being able to get over Hakeem having only two titles.

Okay, you have now done this in multiple threads over multiple days. You are obviously a manifestly unpleasant person, who seems unable to discuss things without resorting to personal attacks, sarcastic remarks, and willful misrepresentations in order to try to ridicule. I am not going to respond to anything you say in the future (I should’ve known better here but I gave it another chance). And I would politely request that you never again respond to anything I say (or make any sarcastic references to my posts while in discussions with others). I don’t want any engagement with you whatsoever, since you are obviously a toxic person, and engaging with (or being engaged by) a toxic person on the internet is really not pleasant.

[LukaTheGOAT]

“This forums dedicated to people thoroughly researching and analysing players at a level (ostensibly) beyond counting up rings or win shares / VORP seems to have a general sense of agreement that this player is really good. Sounds awfully like a case of groupthink.”

But even then, it does not seem you are actually interested in the numbers. To me, when you ask for a metric, and I provide a regular season metric for a high longevity playoff elevator suggesting he had a top ten prime, that would probably indicate a top ten placement. And when other metrics make him look comparable to guys you have a full tier ahead of him, while again having a playoff and/or longevity advantage, someone legitimately curious would probably give that some thought.

Not you though! Guess it really does all come back to ring-counting.

You didn’t provide any numbers except a prime WOWY score that has Hakeem 10th. But, as I imagine you are aware, prime WOWYR has him 49th, and 5-year AuPM has him 61st, while box-score-based metrics are not *crazy* high either. Looking at the overall picture, it is not really of a player who has a metric-based case for being ranked super high, unless there’s metrics I don’t know about, but I asked if there were and you did not provide me with anything I was not already aware of. And not having a great metrics-based case is fine! Metrics are flawed and don’t show the whole picture, and metrics are even more limited for a player whose career was before tracking data since there’s fewer metrics to look at.

Anyways, not sure where you get off saying that “it really does all come back to ring-counting” for me, since I didn’t say anything suggestive of that, and you can look at my personal ranking I put in the thread on the personal list thread and it’s plainly not merely a ring-counting list. I think you’re just wanting to be a bit personal here, which is disappointing.

We don't have PS AuPM for Hakeem in 94 or 95, which much of his argument would be based off of.

Relative to era, Hakeem's peak is up there.

From 93-95, the Rockets were about a 1 point better than the defenses they faced in the regular season. But in 57 playoff games, with Hakeem improving his play, Houston was 5.3 points better than the defenses it faced. The Rockets went from basically around a 50-win level team or so during the season, to posting an SRS comparable to a 62-win pace. (Summarized from Backpicks.com article on Hakeem).

Hakeem's improvement offensively from the RS to the PS, has a ton to do with that., where he perhaps improves more offensively from the RS to the PS than any star player ever

Watch on YouTube

Also:

Hakeem Olajuwon in the 1993 and 1994 postseasons (inflation-adjusted):

➤ 27.6 points/75
➤ 11.8 rebounds/75 (3.0 orb.)
➤ 4.4 assists/75
➤ 1.7 steals/75
➤ 4.2 blocks/75
➤ 57.1% TS% (+4.8 rTS%)
➤ Faced off against -2.4 defenses on average (Just so you know this astronomically high average in terms of the quality of defense)

Also, an older version him showed that he could thrive when giving more talent on the offensive end and it makes me wonder how a younger version of him could potentially look if he got to join up with Barkley and Clyde earlier.

Hakeem has great resiliency against tougher defenses, because of the variety of counters he has on offense and ability to hit tough shots. I think Jokic is a better offensive player, but I just want to highlight this.

For example,

94 Hakeem's Finals might be one of the most dominant performances ever.

Hakeem Olajuwon vs Patrick Ewing in the 1994 Finals (IA per 75):

•) 29.5 points
•) 10.0 rebounds
•) 3.9 assists
•) 1.8 steals
•) 4.3 blocks
•) +5.5 opponent adjusted rTS%
•) Knicks were a -8.1 defense (top 10 all-time)

I also think with a clamped court, that Hakeem's defense is more valuable than even like a Giannis' today for instance. There is less room to cover, and easier for a big to stay by the rim. Hakeem I already thought was a quicker leaper, and generally a better rim-protector because of his timing, but because of the way the game was played, I believe these advantages are amplified. Per Thinking Basketball's video, Hakeem registered at least 5 blocks in 35% of his playoff games from 1986 to 1984...and in 93 and 94, he had at least 5 in nearly HALF of his PS games. This suggests that perhaps 93 and 94 Hakeem isn't quite as far away from his peak on defense as some might believe, but also shows just how much he was able to get his hands on the ball.

A final thing to note about defense during Hakeem's time, is that more offense was ran through the post, which means shut-down 1 v 1 post bigs, also had more value than today.

Like from 88-94, Hakeem held opposing all-star centers 3.6 points per 36 minutes below their season averages, on efficiency with around a 3.5 rTS% drop in efficiency as well per Ben Taylor's GOAT Peaks vid. If you think PS series are won by stars, I could see an argument that Hakeem's defense is proportionally more important in the PS than the RS, because of the drop he can make stars have.

A direct example is

Patrick Ewing vs Hakeem Olajuwon in the 1994 Finals (IA per 75):

•) 19.6 points
•) 12.9 rebounds
•) 1.7 assists
•) 1.3 steals
•) 4.5 blocks
•) -10.7 opponent-adjusted rTS%
•) Rockets were a -4.9 defense (really good)

Ewing’s scoring in the 1994 playoffs before facing Hakeem:

▪️Averaged 23.1 PPG on 54% TS

How Ewing scored in the ‘94 Finals vs Hakeem: Averaged 18.9 PPG on 39%(!) TS

Hakeem shut down an elite center; you usually don't see 1 on 1 defense lead to stars getting locked down, but I think this is one of those special cases where Hakeem's defense was that special.

Ewing-ball leading to so many lost possessions possibly hurt the Knicks more than anything.

[OhayoKD]

Okay diving in here

Is there a statistical basis for the higher views on Hakeem that some people have?

Yes, but it is probably not the sort of stat you're familiar using. And I think it's important to look at methodology here first..

Well, RealGM crashed my post, so this will be kept briefer.

To start, WOWYR is a messy metric that can swing wildly with slight input adjustments or changes in approach, but all the same, Hakeem tends to post values similar to or even better than players like Duncan, Wilt, and Bird — and he has among the absolute best WOWY indicators.

And I think the bottom line is that the case for Hakeem is simply not really an advanced-metric-based case.

To be blunt here, an "advanced metric case" for a player pre-97(not named Jordan) is basically just the box-score as we do not have on/off. Unless you want to go off extremely noisy samples, there is the winning, there is how his team did with him and without him, and there is the box-score.

The "advanced metrics" you're thinking of(setting aside wowyr) are basically box-aggregates for Hakeem. The reason this is important to note is because those aggregates largely use block/steals to ascertain defensive value. Incidentally, when we cross-check box-aggregate scores with real-world impact, we find

A. Those aggregates consistently rate 2-way bigs lower than their impact would suggest(adjusted or "pure")
B. Those aggregates consistently rate higher guards than their real-world impact would suggest(adjusted or "pure")
C. Primary-paint protectors(big or otherwise) do much better by "winning" than they do by the box-score

The on-court reason for this as that smaller players aquire blocks/steals/rebounds often as a byproduct of what their bigger teammates do. There is also no way for a box-aggregate to penalize a bad gamble or to credit deterrence(when the presence of a big or a lebron/pippen esque wing prevents a shot from being attempted). This gets compunded when a stat like BPM actually gives a smaller player more credit for a block than they'd give a bigger player.

I can elaborate more, but the tldr is that when comparing two-way anchors with one-way anchors, the box-score will favor the latter which makes using it as an indication of a player being materially superior at generating wins wierd. It has utility with internal scaling and when you compare similar players, but Hakeem does not generate value the same way Jordan does.

It also so happens that despite being designed to be more stable, even advanced stats *with access to player-tracking and on/off still still score worse in predictivity and stability if they are not directly using an assessment of "winning" as a base(like RAPM).

Simply put, if the goal is to win and you are assessing players as contributors to winning, taking a two-way big and admoninshing them on the basis of RAPTOR(which doesn't even have on/off for hakeem) or BPM or even AUPM(which is BPM mixed with on/off) to an extent, you're probably underrated them.

With that in mind, I'm not sure exactly what you consider "high" but with "winning" Hakeem actually looks pretty comparable as a rs player to the likes of Magic and MJ before we get to the playoffs where
-> Hakeem sees big box-improvement across the board
-> The Rockets as a team see big improvement
-> Hakeem beats a +10psrs team(per sansteere's top 100 teams calc) in his 2nd year(Magic beat one with way more help, Jordan has beaten one)
-> Hakeem's has the 2nd most wins as an srs underdog(Lebron is #1)
-> 2nd best winning percentage as an srs underdog among MVP's(Lebron is #1 again)

Consider Hakeem had the worst on and off court situation of the three by a landslide, consider it took until 92 for him to see his talents maximized schematically, and consider he played the most minutes of the three with no retirements in between, and I think you have a decent case for Hakeem being the best of the three(am actually considering he might have been but I'm not set there).

It also may not be "advanced" but such "real-world data" has certain advantages even if it there we had complete data for alternative impact metrics(we do not)...

-> Sample size. You can get much larger(per-season) samples of off(Ranging up to 82 full-games in a season).
-> Flexibility. You can freely make adjustments and curves as one seens fit. And if you know the direction a supporting cast improves, you can establish an upper-bound or lower-bound on a player's value(example: we set an upper-bound with Jordan by taking his 84 srs and subtracting it from the best regular season score the Bulls posted pre-triangle in 88, we take a lower-bound for Kareem's impact by taking what the Lakers were pre-trade and pretending they didn't lose a bunch of pieces to acquire Kareem)
-> Era-Adjustment. Being worth 7-points in 1969(Russell) is alot better than being worth 7 points in 88 or 2000. Metrics don't really adjust for that atm, you can(a simple way to do this is to track the difference between a player's team and their best possible opponents for a year or a period of years)
-> Real-world impact. Adjusted metrics like RAPM curve down outliers towards the norm. Super-valuable players tend to see their value misattributed to teammates
-> Rotation/Dependency. With real-world data we don't have to worry about wierd lineups or a team being caught off-guard with their best player leaving mid-game. You truly get to see how a team does without a key player.
-> No priors. Winning is noisy, but that noise can cut both ways.

All considered, while I can understand the temptation to dimiss that sort of "stat" as less sophisticated, there are advantages to keeping things "simple" which gets us to...

So there are a great deal of issues with WOWYR....
https://forums.realgm.com/boards/viewtopic.php?p=104353387#p104353387
Think I cover things pretty thoroughly there but, quoting WOWYR's creator, here are they key ones:

In order to accurately solve for “what’s the most likely impact for Larry Bird on all of his lineups?” we need to know about the value of his teammates, like Reggie Lewis. And since Lewis only played a few years, his estimate is a bit fuzzy, and that in turn effects Bird’s estimate

Second, like any RAPM study that’s too long, it smoothes over differences between peak years, ignoring aging and injury. There are some ways around this — one of which is to use smaller time periods

Basically, two of the biggest advantages of WOWY/Indirect go away when we use WOWYR. Instead of 10-82 games from a season, you're working off 2(russell), 8(hakeem) or uh..,TBD with Jordan(as covered, it's never made clear if 94 or 95 are included).
The "adjustments" come from even tinier samples diluting the data further. And while it may be tempting to equate that to the adjustments made with RAPM...

A regression is only as good as its data inputs.

WOWYR regresses RAPM using score differential by game, which is... at least 200 times worse than per-possession data in terms of granularity? (much worse than that actually)

It's an extremely ambitious and fun concept but its value is very limited, as shown by Ben himself... well, not really using it that much. I would just not use WOWYR unless it's for extremely rough classifications (i.e. "good" vs "bad").

EDIT: corrected by quarter -> by game

Incidentally WOWYR'S own creator barely features it in his writeups or videos.

But it’s just a bit surprising to me how high most people on this particular sub-forum seem to be on him, when I see his case as being less metric-based than most.

The forum is more "impact on winning" based. But the type of metrics you are bringing up are weighed less on this board than others specially because they are not well tied to winning. They were more frequently used a year ago, but since that period, I think there's been a decline in how seriously they've been taken, probably in part due to threads like the one I linked above.

History also favors Olajuwon. Going by winning, players who can anchor defenses consistently score as the most valuable and successful players with Russell running the 60's to an extent no player has since. Lebron, Duncan, and KG rule data-ball empirically, and the 70's was all Kareem's.

The most consistent winners were Russell and Duncan. Russell and Kareem experienced the most team success of any MVP.

Russell, Kareem and Lebron are the three biggest historical outliers by discernible "impact" whether you look at peak, prime, post-prime, pre-prime or pre-nba(using Lebron's college-age years for this comparison).

Hakeem is probably the most similar to Russell defensively(possibly excepting Davis) and is also an all-time post scorer.

Based on what usually happens, it would make sense for Hakeem to be a league-best candidate and as it so happens, the largest data samples we do have suggest he was. Especially when it mattered most.

[lessthanjake]

Okay diving in here

Is there a statistical basis for the higher views on Hakeem that some people have?

Yes, but it is probably not the sort of stat you're familiar using. And I think it's important to look at methodology here first..

Well, RealGM crashed my post, so this will be kept briefer.

To start, WOWYR is a messy metric that can swing wildly with slight input adjustments or changes in approach, but all the same, Hakeem tends to post values similar to or even better than players like Duncan, Wilt, and Bird — and he has among the absolute best WOWY indicators.

To be blunt here, an "advanced metric case" for a player pre-97(not named Jordan) is basically just the box-score as we do not have on/off. Unless you want to go off extremely noisy samples, there is the winning, there is how his team did with him and without him, and there is the box-score.

The "advanced metrics" you're thinking of(setting aside wowyr) are basically box-aggregates for Hakeem. The reason this is important to note is because those aggregates largely use block/steals to ascertain defensive value. Incidentally, when we cross-check box-aggregate scores with real-world impact, we find

A. Those aggregates consistently rate 2-way bigs lower than their impact would suggest(adjusted or "pure")
B. Those aggregates consistently rate higher guards than their real-world impact would suggest(adjusted or "pure")
C. Primary-paint protectors(big or otherwise) do much better by "winning" than they do by the box-score

The on-court reason for this as that smaller players aquire blocks/steals/rebounds often as a byproduct of what their bigger teammates do. There is also no way for a box-aggregate to penalize a bad gamble or to credit deterrence(when the presence of a big or a lebon/pippen esque wing prevents a shot from being attempted). This gets compunded when a stat like BPM actually gives a smaller player more credit for a block than they'd give a bigger player.

I can elaborate more, but the tldr is that when comparing two-way anchors with one-way anchors, the box-score will favor the latter which makes using it as an indication of a player being materially superior at generating wins wierd. It has utility with internal scaling and when you compare similar players, but Hakeem does not generate value the same way Jordan does.

It also so happens that despite being designed to be more stable, even advanced stats *with access to player-tracking and on/off still still score worse in predictivity and stability if they are not directly using an assessment of "winning" as a base(like RAPM).

Simply put, if the goal is to win and you are assessing players as contributors to winning, taking a two-way big and admoninshing them on the basis of RAPTOR(which doesn't even have on/off for hakeem) or BPM or even AUPM(which is BPM mixed with on/off) to an extent, you're probably underrated them.

With that in mind, I'm not sure exactly what you consider "high" but with "winning" Hakeem actually looks pretty comparable as a rs player to the likes of Magic and MJ and he's also probably the best riser of the bunch(here we can compare box-production as a tool for internal-scaling, as well as track team-wide improvement or lackthereof):

And then we get to the playoffs where
-> Hakeem sees big box-improvement across the board
-> The Rockets as a team see big improvement
-> Hakeem beats a +10psrs team(per sansteere's top 100 teams calc) in his 2nd year(Magic beat one with way more help, Jordan has beaten one)
-> Hakeem's has the 2nd most wins as an srs underdog(Lebron is #1)
-> 2nd best winning percentage as an srs underdog among MVP's(Lebron is #1 again)


Consider Hakeem had the worst on and off court situation of the three by a landslide, consider it took until 92 for him to see his talents maximized schematically, and consider he played the most minutes of the three with no retirements in between, and I think you have a decent case for Hakeem being the best of the three(am actually considering he might have been but I'm not set there).

It also may not be "advanced" but such "real-world data" has certain advantages even if it there we had compelte data for alternative impact metrics(we do not):

-> Sample size. You can get much larger(per-season) samples of off(Ranging up to 82 full-games in a season).
-> Flexibility. You can freely make adjustments and curves as one seens fit. And if you know the direction a supporting cast improves, you can establish an upper-bound or lower-bound on a player's value(example: we set an upper-bound with Jordan by taking his 84 srs and subtracting it from the best regular season score the Bulls posted pre-triangle in 88, we take a lower-bound for Kareem's impact by taking what the Lakers were pre-trade and pretending they didn't lose a bunch of pieces to acquire Kareem)
-> Era-Adjustment. Being worth 7-points in 1969(Russell) is alot better than being worth 7 points in 88 or 2000. Metrics don't really adjust for that atm, you can(a simple way to do this is to track the difference between a player's team and their best possible opponents for a year or a period of years)
-> Real-world impact. Adjusted metrics like RAPM curve down outliers towards the norm. Super-valuable players tend to see their value misattributed to teammates
-> Rotation/Dependency. With real-world data we don't have to worry about wierd lineups or a team being caught off-guard with their best player leaving mid-game. You truly get to see how a team does without a key player.
-> No priors. Winning is noisy, but that noise can cut both ways.

All considered, while I can understand the temptation to dimiss that sort of "stat" as less sophisticated, there are advantages to keeping things "simple" which gets us to...

So there are a great deal of issues with WOWYR....
https://forums.realgm.com/boards/viewtopic.php?p=104353387#p104353387
Think I cover things pretty thoroughly there but, quoting WOWYR's creator, here are they key ones:



Basically, two of the biggest advantages of WOWY/Indirect go away when we use WOWYR. Instead of 10-82 games from a season, you're working off 2(russell), 8(hakeem) or uh..,TBD with Jordan(as covered, it's never made clear if 94 or 95 are included).
The "adjustments" come from even tinier samples diluting the data further. And while it may be tempting to equate that to the adjustments made with RAPM...
[quote+"Jaivl"]A regression is only as good as its data inputs.

WOWYR regresses RAPM using score differential by game, which is... at least 200 times worse than per-possession data in terms of granularity? (much worse than that actually)

It's an extremely ambitious and fun concept but its value is very limited, as shown by Ben himself... well, not really using it that much. I would just not use WOWYR unless it's for extremely rough classifications (i.e. "good" vs "bad").

EDIT: corrected by quarter -> by game

Incidentally WOWYR'S own creator barely features it in his writeups or videos.

But it’s just a bit surprising to me how high most people on this particular sub-forum seem to be on him, when I see his case as being less metric-based than most.

The forum is more "impact on winning" based. But the type of metrics you are bringing up are weighed less on this board than others specially because they are not well tied to winning. They were more frequently used a year ago, but since that period, I think there's been a decline in how seriously they've been taken, probably in part due to threads like the one I linked above.

I think you misunderstand what I’m getting at. You seem to be suggesting that I’m talking about box-score metrics, so you spend a while explaining why those aren’t great measures, particularly when it comes to measuring defense. And then you’re talking about how people on this forum are more focused on impact metrics. This definitely is misinterpreting what I’m getting at, and I think making an assumption that I’m substantially less sophisticated/knowledgeable about this stuff than I actually am. You’ll notice in my first post on this subject, I specifically said “ there’s not actually a lot of advanced metrics that go far enough back to capture his impact one way or the other.” I’m talking about impact metrics, specifically because I am well aware that that’s what this subforum is more focused on, and I am well aware of why. I’m also well aware of the crippling limitations that box-score-based metrics have at measuring defense. So my point was primarily that there’s not really much of an impact-metric-based case for Hakeem. And that’s for two reasons—both of which I mentioned in my first post on this. One is that there’s just not many impact metrics that go back far enough for Hakeem, because of a lack of tracking data. And the other is that the impact metrics that can go back that far (things like WOWY ratings—which are obviously going to be less precise than possession-by-possession impact metrics, but they’re what we have in pre-tracking-data eras) aren’t actually generally all that high on Hakeem. I did also note once or twice that box-score-based metrics aren’t *super* high on Hakeem either, but that was merely a sidenote because impact metrics are IMO better.

Ultimately, it seems to me that the case for Hakeem is simply not an impact-metric-based case. Maybe there’d be such a case if we had more impact metrics for that era, but overall he doesn’t grade out super highly in what we do have for that era, so I certainly don’t think we can or should assume he would grade out really high in possession-by-possession impact metrics if we did have them. And we certainly shouldn’t build a case based on such an assumption. The case for Hakeem is based more on intuition than that, I think. It’s primarily things like that our eye test evaluates him as having incredibly good defense, his numbers (including offensive numbers) rose a lot in the playoffs, and he ultimately won titles without a super great supporting cast (particularly in 1994). Those are all valid points! My point was just that this is squishier than a lot of other cases—it relies on things such as an eye-test evaluation of his defensive impact (along with inferences about his defensive impact based on his team generally having really good overall defensive efficiency), an intuitive sense of the quality of his supporting casts, and looking at some basic changes in box score numbers for RS—>Playoffs (which is obviously data-based, but we’ve both just agreed that box-score-based data isn’t necessarily all that meaningful). And, while those are all good points that strike me as valid, to me, a squishier case like that has to lead one to be a bit less certain in one’s conclusion than I think a lot of people are acting. That was what I was getting at.

In essence, my point actually boils down to the exact opposite of your assumption that I was trying to focus on box-score-based measures. My point actually boils down to saying “We shouldn’t have such complete certainty in our ranking of a player if we don’t have impact-based metrics to back up the conclusion.”

[AEnigma]

I will always be grateful to Ben for inventing WOWYR to justify my certainty in and removing any squishiness regarding the case for Bill Russell. Looked like an awful lot of groupthink here before that happened.