Retro Player of the Year 1992-93 UPDATE — Hakeem Olajuwon

[homecourtloss]

Generating complete RAPM for a single season (let alone across multiple) requires hundreds of games, not most of a season for one player and a scattering of them for others.

What’s really wild here is that posters who literally didn’t even know what RAPM was a year ago are all of a sudden experts in how to calculate it even though they would not know where to begin to do so. lol

[lessthanjake]

If the partial RAPM are that unreliable because they have oversampled Bulls games, then how come Scottie doesn't dominate the list and only Jordan does..? Pippen is 20th, 22nd, 2nd in RAPM in 1991, 1993 and 1996 respectively.

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I find this quite puzzling as an argument. When a team does really well, we’d fully expect that team to have a lot of players high up in terms of RAPM. That’s how RAPM works (it rewards success when you’re on the court!), especially with no box prior pulling the less productive players down relative to star players.

You can take a look at pure RAPM for the 1996-1997 season, at the website I linked to earlier in this thread. For that season we have full data, and so this issue you raise is not a factor at all. And yet, in the 1996-1997 season, the Bulls have 3 players in the top 10 and 5 players in the top 20. And if you go to 1997-1998, the Bulls have 3 players in the top 20, and 4 players in the top 50. You can look at great teams in general and you’ll find similar things, often with somewhat random role players being ranked highly (for instance, Mario Chalmers was ranked #7 in pure RAPM in the 2012-2013 season, and Shane Battier was ranked #27).

On this point, I notice that you conspicuously didn’t mention how the Bulls did in Squared’s 1987-1988 RAPM, even though Squared does have RAPM for that season and Jordan is #1 while the Bulls had a much larger sample than the vast majority of teams. And, of course, when I look at the 1987-1988 RAPM I see the reason you didn’t mention it: The Bulls only had one other player in the top 75 in Squared’s RAPM that year. And that’s not a surprise, since they were only a 3.76 SRS team that year, so we wouldn’t expect the team as a whole to do as well in RAPM.

I think the overall picture here is a pretty clear one—teams that do really well unsurprisingly have multiple players rated highly by pure RAPM. The Squared RAPM data for the Bulls is consistent with that unremarkable fact. Not only do we not need your theory to explain this, but your theory is inconsistent with the Squared data when it comes to 1987-1988.

lessthanjake also correctly pointed out that the Lakers actually have more sampled games than the Bulls in 1988 and yet Jordan is still ahead of Magic.

He does have a lead...which is much smaller than his general advantage.

Jordan has a 13% lead over Magic in 1988, and a 33% lead over Magic in the overall data. First of all, even if for argument’s sake we attributed that entire difference to this sample thing you’re raising, it’d be clear that this factor could not explain why Jordan’s RAPM lead over Hakeem in the overall data is 76%.

Second of all, we’d probably expect the overall data to be a little better for Jordan compared to Magic than just the 1988 data, since 1988 was probably closer to Magic’s peak than to Jordan’s peak, while the overall data includes other years that we’d probably say the opposite about (most notably 1991). Not surprisingly, if you go to the individual season data for 1991, despite having similar samples (Jordan’s sample is less than 15% bigger), Jordan’s RAPM in 1991 was 60% higher than Magic’s. Needless to say, that’s a far bigger gap than the one in the overall multi-year data. And that’s unsurprising because, as it was peak Jordan and not quite peak Magic, the actual gap between the two players was larger in 1991 than overall. So yeah, I think the difference between Jordan’s lead over Magic in individual seasons and his lead in the full data can basically be explained by actual differences in the size of the gap between the two players. The gap between Jordan and Magic was smaller in 1988 than overall, so Magic was closer that year than he is overall, while vice versa is true in 1991. Your theory is not in any way necessary to explain this. Indeed, your theory isn’t even really consistent with this. For instance, the difference in how much Jordan and Magic were sampled in the multi-year data is way larger than the difference in how much Jordan and Magic were sampled in the 1991 data, and yet Jordan’s lead in the 1991 RAPM is larger.

Or that in 1991, the Lakers and Spurs have almost as many sampled games and yet both Magic and Robinson are well behind Jordan.

Not sure that this theory to discredit Squared2020's RAPM is all that convincing.

Not sure why 1991 Drob is worth mentioning here. Yeah, the advantage vs Magic there. No one is arguing Jordan can't end up looking like the best player in general via RAPM. The question is why we're deciding it matters for a down year in 1993 where 2nd place has half the possessions and almost everyone is working with a third.

This is a valid point. Jordan could theoretically be far more impactful than Hakeem (or Magic) overall, but not have been more impactful than Hakeem in 1993. It’s possible. But I’d note two things:

1. You seem keen for arguments’ purposes to label 1993 a “down year” for Jordan. Yet your primary argument against Jordan (whether in this thread or elsewhere) generally revolves around how the Bulls did in 1994. Do you acknowledge that 1993 was clearly a “down year” for Pippen and Grant? There’s actually *a lot* more indication of that than there is that 1993 was a “down year” for Jordan. To be fair, I think you do sometimes acknowledge it to some extent, but it seems worth flagging. It’s not really clear that 1993 was a down-year for Jordan as much as that his best teammates had a real dip (which they followed up with bounce-back years the next year).

2. More generally, I think we can use broader data to draw inferences about specific years. If Jordan and Hakeem’s RAPM isn’t even close overall, it should make us quite skeptical that Hakeem was more impactful in 1993, even if we think Hakeem was better than normal in 1993. The overall data being so clearly in Jordan’s favor provides a baseline assumption that would have to be overcome. The 1993-specific data is definitely flawed (small sample, especially for Hakeem) but definitely does nothing to overcome that assumption.

[lessthanjake]

Generating complete RAPM for a single season (let alone across multiple) requires hundreds of games, not most of a season for one player and a scattering of them for others.

What’s really wild here is that posters who literally didn’t even know what RAPM was a year ago are all of a sudden experts in how to calculate it even though they would not know where to begin to do so. lol

Lol, if this is referring to me, I was running regressions back when OhayoKD was probably in diapers. You’re barking up the wrong tree there.

[OhayoKD]

If the partial RAPM are that unreliable because they have oversampled Bulls games, then how come Scottie doesn't dominate the list and only Jordan does..? Pippen is 20th, 22nd, 2nd in RAPM in 1991, 1993 and 1996 respectively.

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I find this quite puzzling as an argument. When a team does really well, we’d fully expect that team to have a lot of players high up in terms of RAPM. That’s how RAPM works (it rewards success when you’re on the court!), especially with no box prior pulling the less productive players down relative to star players.

You are puzzled teammates doing well is brought up in response to someone saying a teammate is not doing well?

On this point, I notice that you conspicuously didn’t mention how the Bulls did in Squared’s 1987-1988 RAPM, even though Squared does have RAPM for that season and Jordan is #1 while the Bulls had a much larger sample than the vast majority of teams. And, of course

Probably because Djoker didn't bring up 1988? He has Horace Grant showing up 6th that year. Don't think Djoker's argument works there either.

lessthanjake also correctly pointed out that the Lakers actually have more sampled games than the Bulls in 1988 and yet Jordan is still ahead of Magic.

He does have a lead...which is much smaller than his general advantage.

Jordan has a 13% lead over Magic in 1988, and a 33% lead over Magic in the overall data. First of all, even if for argument’s sake we attributed that entire difference to this sample thing you’re raising, it’d be clear that this factor could not explain why Jordan’s RAPM lead over Hakeem in the overall data is 76%.

Second of all, we’d probably expect the overall data to be a little better for Jordan compared to Magic than just the 1988 data, since 1988 was probably closer to Magic’s peak than to Jordan’s peak

Based on...what?

, while the overall data includes other years that we’d probably say the opposite about (most notably 1991). Not surprisingly, if you go to the individual seasons, despite having similar samples (Jordan’s sample is less than 15% bigger), Jordan’s RAPM in 1991 was 60% higher than Magic’s. Needless to say, that’s a far bigger gap than the one in the overall multi-year data.

Sure. He's still a smaller outlier relative to the rest of the sample as a whole. The basis for the theory is how RAPM usually works as well as this sort of sample distribution as well as squared circle's process being mimicked for existing Lebron RAPM. He saw his numbers significantly jump.

This is a valid point. Jordan could theoretically be far more impactful than Hakeem (or Magic) overall, but not have been more impactful than Hakeem in 1993. It’s possible. But I’d note two things:

1. You seem keen for arguments’ purposes to label 1993 a “down year” for Jordan. Yet your primary argument against Jordan (whether in this thread or elsewhere) generally revolves around how the Bulls did in 1994. Do you acknowledge that 1993 was clearly a “down year” for Pippen and Grant? There’s actually *a lot* more indication of that than there is that 1993 was a “down year” for Jordan. To be fair, I think you do sometimes acknowledge it to some extent, but it seems worth flagging. It’s not really clear that 1993 was a down-year for Jordan as much as that his best teammates had a real dip (which they followed up with bounce-back years the next year).

I've consistently used 1992(Bulls Best SRS) for the sake of peak debates. As you have been shown quotes of repeatedly when claiming this. I even humored a less natural comparison via 1996. No one has claimed peak Jordan had +3 impact or whatever.

2. More generally, I think we can use broader data to draw inferences about specific years. If Jordan and Hakeem’s RAPM isn’t even close overall, it should make us quite skeptical that Hakeem was more impactful in 1993, even if we think Hakeem was better than normal in 1993. The overall data being so clearly in Jordan’s favor provides a baseline assumption that would have to be overcome. The 1993-specific data is definitely flawed (small sample, especially for Hakeem) but definitely does nothing to overcome that assumption.

It doesn't need to. We have actual games to look at lol.

[konr0167]

1. Hakeem

This one is pretty easy. Best season statistically. 2-10 to 40-30 to 53-29 is probably the best impact thing for the decade outside of regular season Drob. Best defender, 2nd best offensive player. Doesn’t have a meltdown in the playoffs and is even better than when he has one of the biggest carry jobs ever next year. Speaks pretty poorly of MVP voters he didn’t win


2. Jordan


If I keep him here, it’s because he won. No really, that’s pretty much it lol. Ewing had less help and won more games in the regular season then completely outplayed him head to head. Everyone talks about the first 3 games. How about game 6 where Ewing scores 26 on 64% true-shooting while Jordan scores 24 on 45%. Every single one of Jordan’s teammates shot much better but Jordan shooting 33% kept chucking, passing on window after window for dumb shot after dumb shot. He went partying when his team was on the brink of elimination, thanks to him, and choked with the season on the line in game 3. He choked again in game 6. But he won, had the “arguably best finals performance ever” vs a bad defense mostly being guarded by 1 or no dudes and it was decided that was the best they’d ever seen.

That’s just incredible marketing.

3. Patrick Ewing

Feel bad putting him 3rd honestly. Probably the 2nd most valuable player in the regular season but finishes 5th because MVP voters think defense is steals and weakside blocks and uncontested rebounds. Outplays the “best player ever” and almost pulls off a major upset outsourcing the best or 2nd scorer ever in the game everything was on the line. I don’t know

4. Barkley

Wins MVP and makes the finals. Bad defender but that didn’t stop the Suns from making the Bulls work.


5. Pippen

Weak regular season but the 2nd best player for the three-peat and probably the best during the most critical series (the Knicks). Massive efficiency drop for Drob this year and his playoffs don’t really move me.

[falcolombardi]

I think of the 90's as an era of very "uncomplete" teams that either lacked star high end stars with their great role players or lacked great role players or "sidekicks" behind their star or star duo

Barkley sixers were a top 5~ish team early on despite their second best player, hersey hawkins, being later anywhere from like 4th-6th best player in a stacked seattke team....which missed an actual barkley level lead offensive option to go the next level of truly great team. Utah added a 3rd above average starter in 96 and it made them a top 2 team for the next 3 years, houston got 2 good role players in thorpe and horry and rockets become a contender (more of that later)

In a league like this a team like the bulls having a low end or borderline mvp-tier player as scottie being their 2nd best and a borderline all star tier player as 3rd in grant was a unusually strong support to a league top 2 lead star in jordan

Barkley, who is not as good as hakeem or jordan by almost any reasonable measure got one of the few comparable supporting casts (albeit with a injury struggling kevin johnson diminishing him a bit as someone who could be in pippen level of impact in my view) this talented but more overlapping roster was able to edge out a slightly better reg season and playoffs performance than teams like knicks and houston who had much more mundane rosters around hakeem and ewing. And edged out a win over the similarly deep with talent seattle who was led by worse/less experienced stars

In my view knicks and houston were overachievers thanks in large part to their 2 way defensive anchors of which hakeem defense and offense performance carried the most impact, who deserve a lot of credit for the work they achieved

Bulls and suns meanwhile probably were not as their roster name sheet or as dominant as previous seasons amd while there are good reasons for that (like bulls 2 title runs + olympics for jordan amd scottie or kevin johnson health declining) is still noteworthy they didnt play much better basketball that teams like knicks or specially houston whose stars lacked the same level of talent around them

Barkley is someone i am low-ish on due to defense in the first place, but jordan and bulls seem to have taken a step back defensively from the prior year, his defense wanes and load increases at the expense of some efficiency, the suns finals is pretty impressive volume.... in a series bulls tied by net rating against a fragile defense without strong rim protection, the knicks series is a tale of 2 halfs about jordan super volume scoring getting ugly when it misses as his 3 first games show how ugly jordan/kobe gunning gets when shots start falling and were a miraculous scoring game from pippen away of a 0-3 series vs a underdog knicks

meanwhile hakeem is just as good as he will be in 94, near absolute peak for himself in both ends which is historical defense and extremely valuable offense

DPOY
1-hakeem, no questions. Goat tier defensive peak for anyone not named bill russel

2- mutombo, seems like he is already providing outlier per minute defensive impact like he will showcase the next year vs seattle

3- ewing prolly no longer in his own defensive prime is still a elite rim protector and overall defender in spite of heavy offensive load

OPOY

1- jordan all time scoring season and overall offensive value, incredible volume in plus efficiency, lots of adequate playmaking off his scoring threat and useful additive skills as off rebounding and good off ball game. Not his best scoring efficiency but still a monster offense year

2- barkley, probably a tad below his sixers years peak as his athletism fades a bit but still the best offensive player in a league leading offense first team

3- reggie miller, the early years of reggie run as one of the league most unheralded offense superstars. A playoff riser that fit on any team amd led better offenses than most would think of a historically unrewarded with individuals accolades player.

POY

1-hakeem, though choice but his level of play this season felt a notch higher than what jordan did in 93.

2- jordan also a great choice for 1st but i care more about jordan defensive step down and lack of truly strong efficiency (the volume goes crazy but game to game vs knicks he almost shot them out of the series and suns are a ideal rival to score in bunchrs against)

3- ewing, did somewhat of a worse version og what hakeem did which is still extremely valuable

4- barkley, was the best player on the league best regular season team, got great wins over teams like seattle and played bulls close

5- pippen for being one of the league best players and the core of bulls defense as well as a big part of their offense. Along with reggie one of the league secret superstars of the 90's who wete not truly understoos how good they were until later

[penbeast0]

Barkley also had top 20 talent around him early on with Doc and Moses but had difficulty meshing his talents and those versions of the Sixers underperformed their talent level. He might be one of those guys like Iverson that just needs to be a one or two man show with everyone else providing 3 and D type support.

[Djoker]

If the partial RAPM are that unreliable because they have oversampled Bulls games, then how come Scottie doesn't dominate the list and only Jordan does..? Pippen is 20th, 22nd, 2nd in RAPM in 1991, 1993 and 1996 respectively.

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I'm sorry but I should have said possessions not games. That argument is that those players with larger numbers of possessions sampled have an edge over those with smaller numbers sampled which really skews the data. Thus it's relevant to look at Pippen who played a high minute load and thus has high possessions to see if his numbers are inflated. They are very clearly not. In fact, RAPM paints him worse than his reputation would suggest.

Including Bulls role players who played lower minutes (and thus fewer possessions) is irrelevant.

lessthanjake also correctly pointed out that the Lakers actually have more sampled games than the Bulls in 1988 and yet Jordan is still ahead of Magic.

He does have a lead...which is much smaller than his general advantage.

What general advantage. In the 1985-1996 multi-year RAPM, Jordan has a large lead over Magic.

Or that in 1991, the Lakers and Spurs have almost as many sampled games and yet both Magic and Robinson are well behind Jordan.

Not sure that this theory to discredit Squared2020's RAPM is all that convincing.

Not sure why 1991 Drob is worth mentioning here. Yeah, the advantage vs Magic there. No one is arguing Jordan can't end up looking like the best player in general via RAPM. The question is why we're deciding it matters for a down year in 1993 where 2nd place has half the possessions and almost everyone is working with a third.

Robinson is one of the impact titans of that era so seems a worthwhile mention. Jordan also leads in RAPM in 1993 which was his "down year" so yea...

You also never addressed this:

Not really sure where to start here.

1. Lebron's WOWY is not the end-all be all. He is also an era-outlier in RAPM despite spending several prime-years staggering with a similar teammate playing a position up, and also the most tested player situationally, and also an outlier in older and younger years, and also a Hakeem-level postseason riser. Jordan has a shot at one of those things. At the moment, Magic has most of them.

2. A season's worth of RAPM is not 82 games, it is a sample of what will overwhelmingly be 5-8 minute snippets of games stitched together with a few full games here and there. And whatever output comes from there is artificially curved towards 0 as an artificial scale is created to approximate adjustments to said 5-8 minute snippets. And in this case it is a sample dominated by one team that like most good teams of the era platooned a bunch. And in this season you are using it against a player who has 1/6th of the sample.

The main benefit of RAPM is it's stabler, but the trade-off of stability is you are no longer tracking real season to season impact, you are tracking a baseline. Even if you want to take that baseline at face-value, the 1-year outputs are artificially fitted to that baseline.

That actual stretches of 82 or near 82 games consistently have Jordan's teammates much better than the bits based on him missing a few minutes a game is a pretty big hint lineup effects are at play here.

I'm very confused where you got the idea that RAPM uses full games of off tbh. Or that the argument for Lebron as an era outlier is exclusively about WOWY

I didn't because respond to that because Lebron is not a subject of this thread and I want to keep the discussion on topic.

Your description of RAPM is fairly accurate and I don't deny any of that. And to stabilize RAPM it definitely takes a certain sample of games.

What I will say is Lebron is NOT an era outlier in RAPM. Prime Curry is in fact better than prime Lebron in terms of RAPM and most of the other impact metrics. Lebron leads when we take a very large sample (ex. 20 years) which reflects positively on his longevity for sure but on year-by-year basis, Lebron is not an outlier.

Going back to Jordan vs. Hakeem which was the original argument and relevant to this thread, Jordan just dominated Hakeem in almost all (or even absolutely all?) the impact stats we have available. Raw ON-OFF, raw playoff ON-OFF, Squared's RAPM, JE Approximation, Moonbeam's data etc. None of them have Hakeem appear as Jordan's equal let alone better.

[Hook_Em]

1. Hakeem

This one is pretty easy. Best season statistically. 2-10 to 40-30 to 53-29 is probably the best impact thing for the decade outside of regular season Drob. Best defender, 2nd best offensive player. Doesn’t have a meltdown in the playoffs and is even better than when he has one of the biggest carry jobs ever next year. Speaks pretty poorly of MVP voters he didn’t win


2. Jordan


If I keep him here, it’s because he won. No really, that’s pretty much it lol. Ewing had less help and won more games in the regular season then completely outplayed him head to head. Everyone talks about the first 3 games. How about game 6 where Ewing scores 26 on 64% true-shooting while Jordan scores 24 on 45%. Every single one of Jordan’s teammates shot much better but Jordan shooting 33% kept chucking, passing on window after window for dumb shot after dumb shot. He went partying when his team was on the brink of elimination, thanks to him, and choked with the season on the line in game 3. He choked again in game 6. But he won, had the “arguably best finals performance ever” vs a bad defense mostly being guarded by 1 or no dudes and it was decided that was the best they’d ever seen.

That’s just incredible marketing.

3. Patrick Ewing

Feel bad putting him 3rd honestly. Probably the 2nd most valuable player in the regular season but finishes 5th because MVP voters think defense is steals and weakside blocks and uncontested rebounds. Outplays the “best player ever” and almost pulls off a major upset outsourcing the best or 2nd scorer ever in the game everything was on the line. I don’t know

4. Barkley

Wins MVP and makes the finals. Bad defender but that didn’t stop the Suns from making the Bulls work.


5. Pippen

Weak regular season but the 2nd best player for the three-peat and probably the best during the most critical series (the Knicks). Massive efficiency drop for Drob this year and his playoffs don’t really move me.

Barkley wasn’t a bad defender in 93’. Id say above average or average at worst which is why he’s no lower than 3rd in this question. He had some insane stat lines in the playoffs and hit clutch shots. Of course game 6 vs the Bulls was his worst so that will stand out.

[One_and_Done]

Barkley also had top 20 talent around him early on with Doc and Moses but had difficulty meshing his talents and those versions of the Sixers underperformed their talent level. He might be one of those guys like Iverson that just needs to be a one or two man show with everyone else providing 3 and D type support.

This is a weird take.

Firstly, Barkley never really had a team of 3&D type support players.

Secondly, his Sixers teams (with the exception of his first few years in the league) were ass. Even when Moses and Doc were there, they weren't at their best anymore.

I don't think there's much evidence for the idea that Barkley was tough to build around or needed hyper special pieces. He just needed to play with some talent. In his first 2 years when he had talent, albeit aging talent, they won 58 and 54 games, and lost to better teams in the playoffs.

[tsherkin]

Barkley also had top 20 talent around him early on with Doc and Moses but had difficulty meshing his talents and those versions of the Sixers underperformed their talent level. He might be one of those guys like Iverson that just needs to be a one or two man show with everyone else providing 3 and D type support.

This is a weird take.

Firstly, Barkley never really had a team of 3&D type support players.

Secondly, his Sixers teams (with the exception of his first few years in the league) were ass. Even when Moses and Doc were there, they weren't at their best anymore.

I don't think there's much evidence for the idea that Barkley was tough to build around or needed hyper special pieces. He just needed to play with some talent. In his first 2 years when he had talent, albeit aging talent, they won 58 and 54 games, and lost to better teams in the playoffs.

Yeah, 58 wins and the 4th-best O in the league in 85. Barkley was playing under 21 mpg. Moses was All-NBA 1st team, 3rd in the MVP vote, led the league in rebounding. Dr J was a 20 ppg scorer on league average efficiency, and then sucked in the ECFs. Moses was pretty bad from the field in that series as well, Boston kind of overwhelmed them.

About the same result a year later, with Barkley recording his first 20+ ppg season and Erving dipping below league average efficiency. Chuck was All-NBA 2nd Team but somehow not an All-Star, heh. Didn't want to give too many to Philly, I guess, and Erving was on the AS team for some reason (aka popularity). Only 6 games from Andrew Toney.

A year later, no Moses, only 60 games from 36 yo Erving and their overall O fell enough enough that their non-elite D couldn't keep up. 45 wins. 52 games from Andrew Toney and only 20 mpg, in the second-last season of his career.

The talent bleed leading into the remainder of Barkley's time in Philly should be pretty evident at this point.

[70sFan]

Player of the Year

1. Hakeem Olajuwon
2. Michael Jordan
3. Patrick Ewing
4. Charles Barkley
5. David Robinson

So, it's all about Hakeem vs Jordan here. I don't think it's controversial take that Hakeem had a better RS. To take Jordan ahead, I think you need to believe that his postseason run was more impressive that Hakeem's. It's not indefensible take, I truly believe that Jordan had an ATG finals series. The problem is that Jordan was extremely underwhelming against the Knicks. It's not a matter of relative underperformance like the year before when he was still really good (just below his standards). Jordan just wasn't good in that series, period - he really struggled against Knicks defense.

In comparison, Hakeem did very well the next season against the Knicks. He wasn't amazing offensively, but his resilient scoring along with the defensive impact put this performance way ahead what Jordan did in 1993 (or even 1992). Of course, this is a completely different season, but I just don't see the same problem with 1993 Olajuwon playoffs some people see. I noticed that some people argue that Hakeem was underwhelming against Sonics, so I made a short post with the access to full game and some clips breaking down his playmaking. Unfortunately, nobody replied to my post. I hoped Hakeem critics could provide some counterpoints For now, I all see is that he should have won against Seattle and I don't understand why.

Ewing was very impressive against the Bulls. I truly believe that he outplayed Jordan h2h and that's something very few players accomplished. I don't think he deserves the 2nd spot though, because of two things:

1. He had a weaker RS.
2. I don't believe Ewing was ever capable of a performance comparable to Jordan's finals.

Barkley is also up there, he was probably 2nd best player in the RS, but I think his defensive limitations lowered his potential. I also think he regressed as a player overall after leaving Philly, reaching higher heights due to incredible supporting cast.

Robinson wasn't at his best this year, but he's a great choice to finish top 5.

Offensive Player of the Year

1. Michael Jordan
2. Charles Barkley
3. Karl Malone

HM: Reggie Miller

Defensive Player of the Year

1. Hakeem Olajuwon
2. Patrick Ewing
3. David Robinson

HM: Dikembe Mutombo

[Djoker]

Player of the Year

1. Hakeem Olajuwon
2. Michael Jordan
3. Patrick Ewing
4. Charles Barkley
5. David Robinson

So, it's all about Hakeem vs Jordan here. I don't think it's controversial take that Hakeem had a better RS. To take Jordan ahead, I think you need to believe that his postseason run was more impressive that Hakeem's. It's not indefensible take, I truly believe that Jordan had an ATG finals series. The problem is that Jordan was extremely underwhelming against the Knicks. It's not a matter of relative underperformance like the year before when he was still really good (just below his standards). Jordan just wasn't good in that series, period - he really struggled against Knicks defense.

In comparison, Hakeem did very well the next season against the Knicks. He wasn't amazing offensively, but his resilient scoring along with the defensive impact put this performance way ahead what Jordan did in 1993 (or even 1992). Of course, this is a completely different season, but I just don't see the same problem with 1993 Olajuwon playoffs some people see. I noticed that some people argue that Hakeem was underwhelming against Sonics, so I made a short post with the access to full game and some clips breaking down his playmaking. Unfortunately, nobody replied to my post. I hoped Hakeem critics could provide some counterpoints For now, I all see is that he should have won against Seattle and I don't understand why.

Ewing was very impressive against the Bulls. I truly believe that he outplayed Jordan h2h and that's something very few players accomplished. I don't think he deserves the 2nd spot though, because of two things:

1. He had a weaker RS.
2. I don't believe Ewing was ever capable of a performance comparable to Jordan's finals.

Barkley is also up there, he was probably 2nd best player in the RS, but I think his defensive limitations lowered his potential. I also think he regressed as a player overall after leaving Philly, reaching higher heights due to incredible supporting cast.

Robinson wasn't at his best this year, but he's a great choice to finish top 5.

Offensive Player of the Year

1. Michael Jordan
2. Charles Barkley
3. Karl Malone

HM: Reggie Miller

Defensive Player of the Year

1. Hakeem Olajuwon
2. Patrick Ewing
3. David Robinson

HM: Dikembe Mutombo

This is a very reasonable and tempered post.

The only thing I'll say is that I think it's unfair to say that Jordan just wasn't good against the Knicks. He put up 32.2/6.2/7.0 on +1.8 rTS and had a very low 2.3 turnovers per game. I feel like if he missed fewer shot in each game but committed one more turnover instead, people would be a lot higher on the series because everyone factors in scoring efficiency but not turnovers into offensive efficiency. Even if we go game-by-game, he had an insanely good Game 4 and a very good Game 5 so it wasn't all bad.

I see this series by Jordan in a similar light to Kareem's series in the 1972 WCF. Not very good but not terrible efficiency at all (still +rTS) and still a very good performance.

[AEnigma]

If the partial RAPM are that unreliable because they have oversampled Bulls games, then how come Scottie doesn't dominate the list and only Jordan does..? Pippen is 20th, 22nd, 2nd in RAPM in 1991, 1993 and 1996 respectively.

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I'm sorry but I should have said possessions not games. That argument is that those players with larger numbers of possessions sampled have an edge over those with smaller numbers sampled which really skews the data. Thus it's relevant to look at Pippen who played a high minute load and thus has high possessions to see if his numbers are inflated. They are very clearly not. In fact, RAPM paints him worse than his reputation would suggest.

Including Bulls role players who played lower minutes (and thus fewer possessions) is irrelevant.

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

lessthanjake also correctly pointed out that the Lakers actually have more sampled games than the Bulls in 1988 and yet Jordan is still ahead of Magic.

He does have a lead...which is much smaller than his general advantage.

What general advantage. In the 1985-1996 multi-year RAPM, Jordan has a large lead over Magic.

While excluding entire years, let alone being grossly incomplete in the years used.

What I will say is Lebron is NOT an era outlier in RAPM. Prime Curry is in fact better than prime Lebron in terms of RAPM and most of the other impact metrics. Lebron leads when we take a very large sample (ex. 20 years) which reflects positively on his longevity for sure but on year-by-year basis, Lebron is not an outlier.

Factually and demonstrably wrong. Maybe it might read as correct if all we had were incomplete RAPM for 2005, 2014, 2015, 2018, and 2022, but fortunately that is not a real standard. At least the gross misuse of Squared’s plus/minus tracking is on-topic and more misleading than outright counterfactual; this is neither.

Going back to Jordan vs. Hakeem which was the original argument and relevant to this thread, Jordan just dominated Hakeem in almost all (or even absolutely all?) the impact stats we have available. Raw ON-OFF, raw playoff ON-OFF, Squared's RAPM, JE Approximation, Moonbeam's data etc. None of them have Hakeem appear as Jordan's equal let alone better.

Telling how the vast majority of that does not show up for both players in the year in question, and the two which kind-of do are by far the least valuable (and then Moonbeam’s is not a single-year measure).

[lessthanjake]

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I find this quite puzzling as an argument. When a team does really well, we’d fully expect that team to have a lot of players high up in terms of RAPM. That’s how RAPM works (it rewards success when you’re on the court!), especially with no box prior pulling the less productive players down relative to star players.

You are puzzled teammates doing well is brought up in response to someone saying a teammate is not doing well?

Pippen has large samples in Squared’s data just like Jordan, and Djoker pointed out that Pippen didn’t do all that well (compared to his own reputation). This would obviously tend to suggest your argument is misguided. You didn’t deny that. Instead, your response was basically just to say that other Bulls players did pretty well, seeming to suggest that being a relatively large portion of the sample clearly helped them. I pointed out that it’s not at all surprising for players to have high RAPMs on a team that did well, so this doesn’t mean those other players had high RAPMs because of the portion of the sample they comprise (and Djoker further pointed out that some of those other Bulls players didn’t have particularly large samples anyways). That just makes the fact that Pippen didn’t generally do all that well *even more damning* for your argument. Pippen having similar-sized samples as Jordan didn’t even lead to him generally having the type of RAPM we’d think a player of his reputation would have, even while on a team that was really good! This certainly strongly suggests that your argument about the relative focus of the sample isn’t particularly valid (or at the very least isn’t nearly as important as you’re seeking to portray).

On this point, I notice that you conspicuously didn’t mention how the Bulls did in Squared’s 1987-1988 RAPM, even though Squared does have RAPM for that season and Jordan is #1 while the Bulls had a much larger sample than the vast majority of teams. And, of course

Probably because Djoker didn't bring up 1988? He has Horace Grant showing up 6th that year. Don't think Djoker's argument works there either.

Djoker didn’t mention Pippen in 1988, because Pippen wasn’t a particularly good player yet in 1988 and he wasn’t a starter. It’s not a year that fits with the specific point Djoker was making, since Djoker’s point was about Pippen making up a similar portion of the sample as Jordan (which isn’t true in 1988). But it is still a year where the Bulls were sampled more than virtually every other team (and therefore is a year where your theory should be applicable), and yet we see that the Bulls didn’t do particularly well in RAPM. This is not surprising, since the team didn’t do super great as a whole, but it isn’t consistent with your argument.

You point out that Horace Grant was 6th in the 1988 RAPM, but Grant doesn’t even have that high a sample there (he was a rookie that wasn’t starting that year), nor was he particularly good that year. Him being high up is likely just a function of collinearity or random variance, rather than being a large portion of the sample (which he wasn’t). Noting that a Bulls player with a much smaller sample than Jordan’s did well (and, indeed, better than most players in the sample that had larger sample sizes) doesn’t actually do much of anything to support your argument.

To test a point in 1988 that’s similar to what Djoker was making with Pippen, we’d need to look at players with similar samples to Jordan and see what we find. The Bulls player with a similar sample to Jordan that year was Charles Oakley, who can be found all the way down at 94th. So that doesn’t support your point. But, of course, you could validly point out that Charles Oakley just isn’t a particularly good player, so no model could ever get confidence to move him far from the mean. And that’s probably right. Oakley wasn’t good (despite arguments you’ve tried to fashion to the contrary in the past). However, we actually have a couple other teams that were sampled even more than the Bulls that year—specifically the Lakers and Celtics. And yet no one but Magic Johnson is even in the same stratosphere as Jordan. And that includes Larry Bird with a virtually identical sample as Jordan being all the way down at 71st. I know you’re not as high on Bird as most people, but I think we can both agree that Bird being in 71st in 1988—a prime year where he finished 2nd in MVP voting—is definitely far lower than where we’d ever reasonably expect. Of course, this is almost certainly largely a result of random variance, but if the portion-of-the-sample issue you’re trying to raise actually had some kind of significant effect, we wouldn’t expect to see a result like this.

He does have a lead...which is much smaller than his general advantage.

Jordan has a 13% lead over Magic in 1988, and a 33% lead over Magic in the overall data. First of all, even if for argument’s sake we attributed that entire difference to this sample thing you’re raising, it’d be clear that this factor could not explain why Jordan’s RAPM lead over Hakeem in the overall data is 76%.

Second of all, we’d probably expect the overall data to be a little better for Jordan compared to Magic than just the 1988 data, since 1988 was probably closer to Magic’s peak than to Jordan’s peak

Based on...what?

I think reasonable minds can differ on how close 1988 was to Jordan’s peak, so I won’t bother going down a rabbit hole on that. However, I think most anyone would generally expect Jordan to have a bigger lead over Magic in 1991 than in 1988—which is exactly what we see in the Squared data. I really don’t think it’s all that surprising for Jordan to have a modestly smaller lead over Magic in 1988 than he does in the overall multi-year data, which includes some more favorable years for Jordan. Heck, a small portion of Magic’s data in the multi-year sample comes from 1996. I believe it’s only like 8% of the sample for Magic, so it can’t have made a large difference, but that fact by itself would also at least contribute to an expectation that Jordan’s lead in the overall data would be a little higher than his lead in 1988. None of this is surprising. Not to mention that there’s also random variance that makes relatively minor shifts in RAPM unsurprising in general. You’re trying to act like your theory being correct is the only possible explanation for something that really isn’t surprising at all. The reality is that there’s not really circumstantial evidence for your theory (in fact, it’s the opposite), beyond the fact that Jordan is dominant in Squared’s RAPM. But of course the simplest explanation for that is that the consensus GOAT just was that good.

, while the overall data includes other years that we’d probably say the opposite about (most notably 1991). Not surprisingly, if you go to the individual seasons, despite having similar samples (Jordan’s sample is less than 15% bigger), Jordan’s RAPM in 1991 was 60% higher than Magic’s. Needless to say, that’s a far bigger gap than the one in the overall multi-year data.

Sure. He's still a smaller outlier relative to the rest of the sample as a whole. The basis for the theory is how RAPM usually works as well as this sort of sample distribution as well as squared circle's process being mimicked for existing Lebron RAPM. He saw his numbers significantly jump.

I have no idea what you mean about “as well as squared circle’s process being mimicked for existing Lebron RAPM” and I also think making a seemingly sarcastic pun out of Squared’s username is in bad taste.

More on topic, what relevant point do you even think you’re making here? We were discussing the gap between Jordan and Magic in the overall multi-year data compared to the gap between them in the 1988 data. The reason we were discussing this is because you were trying to assert that the reason there’s a bigger gap between them in the overall data than in 1988 is because Jordan’s portion of the sample compared to Magic’s is higher in the overall data than in 1988. The reason the 1991-specific data was brought up here is because the fact that there’s a bigger gap between them in the multi-year data than in 1988 is obviously in part a function of the fact that there’s a large gap between them in the 1991 data. Your point here doesn’t map onto that discussion in any meaningful way.

Nor does it map onto the discussion about your theory for discounting Jordan’s dominance in Squared RAPM more generally. Who cares whether Jordan is “a smaller outlier relative to the rest of the sample as a whole” in 1988 or 1991? How does that map onto your argument? Bulls games make up 22.8% of the 1988 sample and 19.4% of the 1991 sample. Hardly a major difference that would lead us to expect any significant difference in the two years based on your theory (nor do we see any such significant difference, anyways, so you wouldn’t have much of any point here even if the portions of the sample differed a lot). More generally, it may be a little surprising for Jordan to arguably be a bigger outlier in 1988 than in 1991 (though that’s arguable, since the gap to #2 is larger in 1991), but we should remember that the 1991 Bulls sample is highly skewed against the Bulls (they absolutely dominated the unsampled games), so the 1991 data probably underrates what Jordan would look like with a full sample. Regardless of any of this, though, it is certainly objectively the case that the gap between Jordan and Magic is bigger in 1991 than in 1988, and that the 1991 data clearly pulls up the gap in the multi-year data.

[falcolombardi]

Lebrob being a outlier in 20 year RAPM is probably more impressive than anythingh in smaller sample precisely because it showcases his massive year after year value add and that his average is higher than anyone else -despite- a longer sample

If two guys shot well from 3 but one shots twice as much he is more impresive, lebron having the highest career average despite the more time for down years to happen and playing from such a young age only makes it more crazy, not less lol

But even then is not true that lebron is not a outlier in smaller "peak" sample of plus-minus metrics lol

https://www.thespax.com/nba/quantifying-the-nbas-greatest-five-year-peaks-since-1997/

"LeBron has two unique five year stretches during which he had a higher RAPM than all but one other player’s best single five year stretch. In other word, LeBron’s 6.15 RAPM from 2006 to 2010 is the 4th highest five year RAPM peak since 1997. Of the three five year stretches ahead, two belong to LeBron and both of those stretches didn’t overlap with the 2006-10 stretch at all."

He also has the best career playoff rapm here in the biggest sample size to boot

https://public.tableau.com/views/PostseasonRAPM1997-2021/PostseasonRAPM1997-2021?%3Aembed=y&%3AshowVizHome=no#2

I respect squared effort but a tiny and unbalanced sample of single seasons is not a reliable metric lmao, much less when we have actual full data for other players

[lessthanjake]

Why are you only looking at Pippen?

The Bulls have 3 of the top 20 and 4 of the top 30 in 1991.

The Bulls have 3 of the top 30 and 4 of the top 50 in 1993

The Bulls have the top 2, 3 of the top 10, and 4 of the top 20 in 1996

I'm sorry but I should have said possessions not games. That argument is that those players with larger numbers of possessions sampled have an edge over those with smaller numbers sampled which really skews the data. Thus it's relevant to look at Pippen who played a high minute load and thus has high possessions to see if his numbers are inflated. They are very clearly not. In fact, RAPM paints him worse than his reputation would suggest.

Including Bulls role players who played lower minutes (and thus fewer possessions) is irrelevant.

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

I’d be careful assuming the sample “transparently misses three of [Magic’s] best years.” It seems quite clear that the multi-year data includes more data than what Squared has posted on his website for individual seasons. Most obviously, the most recent multi-year data was posted specifically mentioning that “[m]ore 1986-87 data” was being added. There is no specific page for 1986-87 RAPM on Squared’s website, so obviously that’s an example of data that’s in the multi-year data but isn’t independently posted on Squared’s website. And that’s a particularly notable example here, since 1986-87 is surely a year you’re including as one of Magic’s best years that Squared purportedly has “transparently misse[d].” I doubt there’s a lot there for Magic, since Squared said he only logged about 100 games from 1986-87, but it is likely that the sample does actually include some additional Magic data from at least 1986-87 (and possibly other years).

I’ll also note that it’d be nice if we could ask Squared about exactly what is in this multi-year data and not on the website (and potentially ask that individual-season stuff be put on the website), but that is now impossible for reasons you should be well aware of.

[falcolombardi]

I'm sorry but I should have said possessions not games. That argument is that those players with larger numbers of possessions sampled have an edge over those with smaller numbers sampled which really skews the data. Thus it's relevant to look at Pippen who played a high minute load and thus has high possessions to see if his numbers are inflated. They are very clearly not. In fact, RAPM paints him worse than his reputation would suggest.

Including Bulls role players who played lower minutes (and thus fewer possessions) is irrelevant.

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

I’d be careful assuming the sample “transparently misses three of [Magic’s] best years.” It seems quite clear that the multi-year data includes more data than what Squared has posted on his website for individual seasons. Most obviously, the most recent multi-year data was posted specifically mentioning that “[m]ore 1986-87 data” was being added. There is no specific page for 1986-87 RAPM on Squared’s website, so obviously that’s an example of data that’s in the multi-year data but isn’t independently posted on Squared’s website. And that’s a particularly notable example here, since 1986-87 is surely a year you’re including as one of Magic’s best years that Squared purportedly has “transparently misse[d].” I doubt there’s a lot there for Magic, since Squared said he only logged about 100 games from 1986-87, but it is likely that the sample does actually include some additional Magic data from at least 1986-87 (and possibly other years).

I’ll also note that it’d be nice if we could ask Squared about exactly what is in this multi-year data and not on the website (and potentially ask that individual-season stuff be put on the website), but that is now impossible for reasons you should be well aware of.

We all would be thankful if you didnt involve someone's terrible illness as part of a basketball discussion

[AEnigma]

I'm sorry but I should have said possessions not games. That argument is that those players with larger numbers of possessions sampled have an edge over those with smaller numbers sampled which really skews the data. Thus it's relevant to look at Pippen who played a high minute load and thus has high possessions to see if his numbers are inflated. They are very clearly not. In fact, RAPM paints him worse than his reputation would suggest.

Including Bulls role players who played lower minutes (and thus fewer possessions) is irrelevant.

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

I’d be careful assuming the sample “transparently misses three of [Magic’s] best years.”

I would be careful reading before assuming a name not mentioned was at play (rather than the one name which was mentioned). Although regardless of the name, the idea of adding a few more scattered games hardly changes much of anything.

I’ll also note that it’d be nice if we could ask Squared about exactly what is in this multi-year data and not on the website (and potentially ask that individual-season stuff be put on the website), but that is now impossible for reasons you should be well aware of.

What reasons are those. Last I checked (summer), he was plenty available and responsive on Twitter for anyone who cared enough to talk with him about the data rather than use him as a prop.

[lessthanjake]

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

I’d be careful assuming the sample “transparently misses three of [Magic’s] best years.” It seems quite clear that the multi-year data includes more data than what Squared has posted on his website for individual seasons. Most obviously, the most recent multi-year data was posted specifically mentioning that “[m]ore 1986-87 data” was being added. There is no specific page for 1986-87 RAPM on Squared’s website, so obviously that’s an example of data that’s in the multi-year data but isn’t independently posted on Squared’s website. And that’s a particularly notable example here, since 1986-87 is surely a year you’re including as one of Magic’s best years that Squared purportedly has “transparently misse[d].” I doubt there’s a lot there for Magic, since Squared said he only logged about 100 games from 1986-87, but it is likely that the sample does actually include some additional Magic data from at least 1986-87 (and possibly other years).

I’ll also note that it’d be nice if we could ask Squared about exactly what is in this multi-year data and not on the website (and potentially ask that individual-season stuff be put on the website), but that is now impossible for reasons you should be well aware of.

We all would be thankful if you didnt involve someone's terrible illness as part of a basketball discussion

I’m not even entirely sure *exactly* what you’re referring to here, to be honest. Perhaps you’re not actually aware of this, but as I recall, the reason Squared doesn’t post here is because he decided to stop posting here as a result of an acrimonious discussion with AEnigma (and eventually others, I believe, but I can’t recall exactly who)—which obviously is the basic fact I was alluding to. It is highly unfortunate, regardless of the substance of that acrimonious discussion (which I don’t even remember much of the details of, to be honest). It *really* would be nice to be able to discuss things with Squared here in situations like this where the discussion is going into detail about his work.

And, I don’t want to be too negative, but I do think it’s generally the case that posters who tend to be unpleasant and aggressive will have a chilling effect on other posters—with this just perhaps being one example of that in play, at least in part. Speaking personally, I’m not Squared and I don’t have any work of mine that others regularly discuss here, but I’ve certainly considered not posting here anymore (and have taken significant breaks from doing so) as a result of unpleasant behavior of the same posters that I believe Squared’s acrimonious discussion occurred with. I don’t think me not posting here would be a big deal (Squared definitely has more to offer than I do, IMO), but it’s all just something to think about for everyone, including AEnigma (and also perhaps including myself, to be fair). But that’s all a bit of a rabbit hole, and I think the discussion can move on from it. Perhaps I shouldn’t have alluded to that incident, because I should’ve known it might open up a can of worms.

The sample transparently misses three of his best years, which is yet another reason these citations to hodgepodge RAPM are clearly not meant to advance any real discussion about the players.

I’d be careful assuming the sample “transparently misses three of [Magic’s] best years.”

I would be careful reading before assuming a name not mentioned was at play (rather than the one name which was mentioned).

Fair enough. Got my wires crossed since I’d been talking so much about Magic in the previous post I’d just written. That said, it wasn’t a totally worthless misunderstanding IMO, since I do think the general fact that the multi-year data includes more than the website has for single-year data is worth people being aware of.

I’ll also note that it’d be nice if we could ask Squared about exactly what is in this multi-year data and not on the website (and potentially ask that individual-season stuff be put on the website), but that is now impossible for reasons you should be well aware of.

What reasons are those. Last I checked (summer), he was plenty available and responsive on Twitter for anyone who cared enough to talk with him about the data rather than use him as a prop.

He’s actually not all *that* responsive on Twitter from what I’ve seen, and obviously there’s no way to get him to address specifics about his work being brought up in a discussion like this one. I actually would be curious his thoughts on many of the things being discussed here, since obviously he’s the primary expert in his own work and the discussion is going into great detail about it. But I don’t see how it’d be possible to get his thoughts, since there is about a 0% chance he’d delve into something relating to discussions here. And that is obviously unfortunate. EDIT: It is what it is, though, so let’s maybe just leave it there—I probably shouldn’t have even alluded to that incident or be elaborating more about it in this post, since it surely wouldn’t be productive to delve back into that episode.