Retro Player of the Year 1995-96 UPDATE — Michael Jordan

[OhayoKD]

Where are you getting 64-win pace with Hakeem and Barkley?

From one of the Fatal9 posts quoted. They were using record not SRS I think. Seems they got "lucky", or were "clutch".

Yes they were lucky. 57-win pace with a big three is not impressive whatsoever. That is context surrounding Hakeem and the Rockets in 1997.

[/quote]
It's a big two but sure. Hakeem's got no RS case.

Sure it's super small off sample but you usually stick with it and WOWY data isn't really telling us that Hakeem is an impact giant.

Sure, in 97 it says that. In 96 over a much larger sample of 10 games the Rockets go 1-9 and are -8 which would indeed paint Hakeem as an impact monster and a MVPish regular-season player if we combine the two samples. I don't know what the net-rating for the Barkley-less games were but at 45-wins that would be a great regular-season.

1996 Hakeem was better in the RS than 1997 Hakeem.

Not particularly relevant to how the Rockets did without him but that sample seems pretty useless so whatever. I don't know why I didn't think to look at later years but we have a larger sample with a similar roster for 98 where they're 14-20. Someone can vet how clean that is, but i'll be treating the rockets as an averagish team without Hakeem as opposed to a bad one as things stand. Maybe a little above given Bird's WOWY stuff

And then Hakeem starts shooting 8 points better in the postseason, increases his assists lowers his turnover economy and for my money ramps up his defense too. If Jordan can get POY votes making 25-win teams into 50 or 45-win ones over prime Magic, then I don't see why Hakeem, with clearer elevation, facing a much weaker competition wouldn't merit POY consideration here

Because Jordan elevated the 1988 and 1989 Bulls beyond what was expected while Hakeem elevated the 1997 Rockets less than what was expected. Jordan in those years didn't have two all-stars on his team.

The 45-win being used here is specifically without Barkley. Not sure why you're arguing vs results with perception. With Barkley they were alot better than Chicago anyway.

Jordan's true-shooting dropped by 4-points despite his volume also dropping by 2 ontop of a slightly higher tov average. Seems like there's reason to believe Jordan got significantly worse while Hakeem got alot better.

Jordan's volume didn't drop 2 points. He went from averaging 29.6 ppg in the RS to 31.1 ppg in the PS so his volume actually went up. The turnovers did go up in the PS from 2.0 to 2.7 but MJ's turnovers were still insanely low for his offensive load and factor into why he was overall a very efficient offensive player. It's very odd to bring up turnovers with MJ.

You are correct, I was looking at per 36 minutes. You aren't looking at his offensive load, you're looking at how often he was the last or second last to touch the ball. Very different things.

Also worth noting that MJ faced a very strong slate of defenses in the 1997 PS which can explain a 4-point efficiency drop.

Bullets: -1.8 rDRtg
Hawks: -4.4 rDRtg
Heat: -6.1 rDRtg
Jazz: -2.7 rDRtg

Also Hakeem didn't elevate his D in the PS at all. In fact, he probably dropped off on that end. In the RS, the Rockets were a -2.7 rDRtg team and in the PS a +0.3 rDRtg team.

Okay but then they went
+16
+10
and
+5

with Hakeem going from just being their lead scorer to being their most efficient scorer by far

You're disappointed in them for being +5 but then they go

+9 PSRS
+8 PSRS
+6 PSRS

with Hakeem going off while the other two wilted in the most important one.

That being said...

1997 Chicago from +10 went to

+8 PSRS
+13 PSRS
+14 PSRS
+8 PSRS

Considering the issues with the 96 sample, Hakeem having a bad WOWY in 97(not so bad if 98 is reasonable for use but still), the Bulls improving in the playoffs as well, and I think I have to agree Jordan should be POY for 97 or at least over Hakeem even though my eyetest liked Hakeem alot better in the POs, especially the Jazz.

I do think Hakeem was definitely better vs Utah, but I've been pretty consistent about letting the data outweigh my priors for POY purposes and the case Hakeem clearly outplayed Jordan in the playoffs isn't very strong empirically.

Still, the Rockets are very good in the postseason and Hakeem is clearly their best player on both ends so I'd probably put him 2. It's also possible i'm overindexing on how both played vs the Jazz. I didn't end up putting Ewing ahead of MJ for outplaying him in 1993, so it would be wierd to put Hakeem over MJ for winning an indirect comparison for one series. There's also actual full RAPM which i'm pretty sure favors Jordan for 97-98 no matter what.

Hakeem never faced more defensive attention than Jordan. Get out of here with that. In the first two rounds, Barkley and Drexler were playing like all-stars.

He definitely did vs Utah.

20 ppg is very low volume. That's not offensive carrying LOL.

He definitely carried vs the Jazz, but you could be right on the first two rounds.

Oh and MJ happens to be not just in front of Hakeem but 1st in RAPM,

I don't doubt he's ahead but name your source?

First sourced set I found comes from "A Screaming Comes Across the Court" on google has MJ 2nd
https://ascreamingcomesacrossthecourt.blogspot.com/2013/10/introducing-1990s-rapm.html?m=1

RPM, RAPTOR for the season too.

And Hakeem is ahead in IBM's metric. If you are not going to justify the relevance of the inputs, I am not going to care about the outputs.

You not caring doesn't make the data magically not exist.[/quote]
It exists as much as the IBM thing exists. Why are we relirigiating this convo from the 1988 thread?

Spoiler:

They're the same type of data. Humans choose what to count and then put weights on what they've counted, That decades were spent enshrining a narrow set of approaches as objectively valuable does not magically give the formulas and inputs you prefer inherent value and pretending it does would get you discredited in any space with an ounce of serious academic rigor.

Beyond the extent you can justify the approach or weightings vs approaches/weightings that favor alternative players, your formulas are not legitimate evidence.

IBM of course is not a few games, Lebronny's tracking covers multiple years of full playoff runs. If sample size is the issue, then the solution is to increase the sample, not keep reinforcing a set of priors that have never been seriously tested because they produce outputs you find convenient.

On/off is even in the regular-season and then it collapses for Jordan. And Magic is clearly advantaged in WOWY (something you cleverly side-stepped by throwing in games Magic did not play with his "with")

I don't care and it seems the voters don't either, for good reason.

Here are the sources. For RAPTOR, you have to download the Excel file.

RPM: https://web.archive.org/web/20230320184333/http://www.espn.com/nba/statistics/rpm/_/year/1997

RAPTOR: https://github.com/fivethirtyeight/data/tree/master/nba-raptor

I asked for RAPM. Not made-up blah blah blah.

[AEnigma]

Votes are tallied. I recorded 13 approved votes: Djoker, AEnigma, B-Mitch 30, ShaqAttac, ILikeShaiGuys, OhayoKD (submitting “kola’s” ballot as his official one), LA Bird, penbeast0, capfan33, konr0167, Paulluxx, One_and_Done, and trelos. DJoker, AEnigma, B-Mitch 30, trelos, OhayoKD, and ILikeShaiGuys voted for both Offensive and Defensive Player of the Year, and capfan33 voted for Defensive Player of the Year. Please let me know if I seem to have missed or otherwise improperly recorded a vote.

1995-96 Results

(Retro) Offensive Player of the Year — Michael Jordan (3)

Code: Select all

Player       1st   2nd   3rd   Points  Shares
1. Michael Jordan   5   1   0    28    0.933
2. Penny Hardaway   1   3   0    14    0.567
3. Shaquille O’Neal   0   1   2    5    0.167
4. Karl Malone   0   0   2    2    0.067
5. Terrell Brandon   0   0   1    1    0.033
5. Vlade Divac  0   0   1    1    0.033

(Retro) Defensive Player of the Year — Scottie Pippen

Code: Select all

Player         1st   2nd   3rd   Points  Shares
1. Scottie Pippen   4   0   0    20    0.571
2. Patrick Ewing    1   4   1    18    0.514
3. David Robinson   2   1   3    16    0.457
4. Hakeem Olajuwon    0   1   0    3    0.086
4. Dale Davis    0   1   0    3    0.086
6. Dennis Rodman    0   0   2    2    0.057
7. Alonzo Mourning    0   0   1    1    0.029

Retro Player of the Year — Michael Jordan (4)

Code: Select all

Player      1st 2nd 3rd 4th 5th Pts  POY Shares
1. Michael Jordan  11  2  0  0  0   124   0.954
2. David Robinson  0  5  3  3  1   60   0.462
3. Penny Hardaway   0  3  3  2  1   43   0.331
4. Shaquille O’Neal  2  0  0  4  0    32   0.246
5. Karl Malone  0  2  2  1  3   30   0.231
6. Hakeem Olajuwon   0  0  4  1  0   23   0.177
7. Gary Payton   0  1  0  1  1   11   0.085
8a. Scottie Pippen  0  0  0  0  6   6    0.046
8b. Patrick Ewing   0  0  1  0  1   6   0.046
10. Alonzo Mourning   0  0  0  1  0   3   0.023

In the prior project, there were 24 votes, with no overlap. These are the aggregated results of the two projects across 37 total ballots:

Spoiler:

Code: Select all

Player   1st 2nd 3rd 4th 5th Pts  POY Shares
1. Michael Jordan  35  2  0  0  0   364   0.984
2. David Robinson  0  16  13  6  1   196   0.530
3. Karl Malone  0  14  6  4  8   148   0.400
4. Penny Hardaway   0  3  8  10  6   97   0.262
5. Gary Payton   0  1  2  9  7   51   0.138
6. Hakeem Olajuwon   0  1  6  3  3   49   0.132
7. Shaquille O’Neal   2  0  0  4  0   32   0.086
8. Scottie Pippen  0  0  1  0  10   15    0.041
9. Patrick Ewing   0  0  1  0  1   6   0.016
10. Alonzo Mourning   0  0  0  1  0   3   0.008
11. Shawn Kemp   0  0  0  0  1   1   0.003

1997 thread is open.

[lessthanjake]

Let's extend this

They were

1-9 and -8 in without Hakeem in 1996 and 3-7 and -4.4 without him in 1995.

I don't know how to get net-ratings for multiple players missing or being included but for 95-97 that means, by record, Houston played at a 23-win pace. Houston going 45-wins with no Barkley seems like good rs impact to me. Not going to use the big three or hakeem/barkley samples for obvious reasons.

Why would you ever “extend” the WOWY analysis here to the games Hakeem missed in 1995-96?

Same reason I extended it for games Hakeem missed in 91 when looking at 93 Hakeem. (that extension hurt him).

Out of the 12 players on the 1996-97 Rockets that played the most minutes besides Hakeem, only 3 of those players even played a single minute for the Rockets in the 10 games that Hakeem missed in 1995-1996. Those were Drexler, Elie, and Sam Mack. And Drexler and Elie both barely played in that span themselves, with Drexler playing a grand total of 92 minutes and Elie playing 74. The only guy that actually played significant minutes on the Rockets during the 10-game span Hakeem missed in 1995-1996 was Sam Mack (who was averaging almost twice as many minutes in that 1995-1996 span than he did in 1996-1997, because of how short-handed the Rockets were in that 1995-1996 span). And that’s not even mentioning that many of the top guys that were on the 1995-1996 Rockets but not on the 1996-1997 Rockets missed a lot of time in that 1995-1996 span, including Horry missing a few games, Cassell barely playing, etc. They were super short-handed in those games and just not at all similar to the 1996-1997 team.

Interesting. I'd agree that's not worth counting then. Seems 1999 Bulls-ish. I'm confused how drexler has 92 minutes though.
https://www.statmuse.com/nba/ask/rockets-record-with-drexler-and-without-hakeem-in-1995
Statmuse says Drexler played 1 game without him in 96

statmuse does say Drexler played 9 of the 10 games missed in 95 though I guess it would be nice if someone could fact check that. Shaving off a loss improves them to -3.4 and 3-6

I think sometimes Statsmuse spits out incorrect information due to not understanding what is being asked.

The 10 games Hakeem missed in the 1995-1996 season were on 3/19, 3/21, 3/26, 3/27, 3/30, 4/2, 4/3, 4/5, 4/7, and 4/16. If you go through the box scores for those games (can start here, though the absences aren’t all in a row: https://www.basketball-reference.com/boxscores/199603190HOU.html), you’ll see all the minutes for the players in each game. Drexler only played 29 minutes in the 4/5 game, 35 minutes in the 4/7 game, and 28 minutes in the 4/16 game.

And, as I mentioned, in general when you go through those box scores you’ll see that the Rockets were horribly short-handed in those 10 games—even beyond Hakeem being out and Drexler barely playing. Elie and Cassell both also barely played in those games, Horry missed a few games, and Kenny Smith missed one as well. Their rotation in those games was mostly Sam Mack, Chucky Brown, Mark Bryant, Tracy Moore, Eldridge Recasner, Melvin Booker, Tim Breaux, Pete Chilcutt, and Charles Jones. It’s definitely no surprise they went 1-9, and that was about a lot more than just Hakeem missing those games. More importantly for purposes of this discussion, it definitely doesn’t allow for any valid WOWY comparison regarding Hakeem in 1996-97—though I see you acknowledge that, so we aren’t in disagreement there.

As for 1994-95, I think it’s also pretty useless as a WOWY data point for 1996-97. You are right that Drexler played in 9 of the 10 games Hakeem missed in 1994-95, but the teams were otherwise almost entirely different. The 1996-97 Rockets had 15 players that played at least 100 minutes, and besides Hakeem only two of them played on the 1994-95 Rockets: Drexler and Mario Elie (the latter of whom was a bench player in 1994-95). I hardly think it’s meaningful how the Rockets did in a different season with Drexler, Elie, and a completely different cast of players. That might as well just be comparing the 1997 Rockets to a different franchise’s record in some other year, because the teams just are so different. In fact, it’s no more similar to the 1997 Rockets than the 51-win 1993 Blazers were (the 1993 Blazers also had Drexler, Elie, and no other overlap with the 1997 Rockets).