Some Historical Plus-Minus

[lessthanjake]

I’ve not done the work to convert the Magic Johnson stuff (I did the Jordan stuff, but it actually takes a while). But eyeballing it, I’m thinking Magic was probably hovering around +20 on-off on average in those samples (probably slightly more in 1985 and slightly less in 1991). Very impressive stuff from him.

[rk2023]

I ran through 1988 and 1991 Magic at a higher level (didn't adjust for pace or minutes yet, and filtered out the partial games).

1988 (48 Games):
8.38 +/- Per-Game with Magic, -2.36 +/- Per-Game without

1991 (47 Games):
5.38 +/- Per-Game with Magic, -1.83 +/- Per-Game without

The former is more in-line with the full 82 MOV/game Lakers results, FWIW.

Interesting. Can you convert to per 48?

Am working on it or at least plan to this evening; realized this formula was more of a 50-50 minute assumption. Plan to adjust for possessions afterwards, then later dive into opponent strength and so on for data visualization.

[Squared2020]

.

[lessthanjake]

For the 1990-1991 games, taking the MOV with Robinson on and the number of possessions with him on (adding the offensive and defensive possessions and dividing by 2), it seems like the Spurs were +5.11 per 100 possessions with Robinson on in that sample.

I also estimated the number of “off” possessions by taking the number of games, the Spurs’s pace that season, and the number of OT periods (it’d be better to do this using the pace of each of these specific games, but I’m a bit lazy for that at the moment). This allowed me to derive an estimated -8.64 “off” net rating. Which would put Robinson’s on-off in this 1990-1991 sample at an estimated +13.75.

[rk2023]

Noticed an error in my earlier methodology of extrapolating the season average of MPG. I fixed such in my calculation for 1991 Jordan (have yet to go through the rest of his, Magic, and Robinson’s numbers).

What I got:

Code: Select all

Net /100 Rating “on-court”: 9.77 (111.96-102.19)
+/- per 48 min, “on-court”: 9.84
Bulls MOV per 48 min: 6.01
+/- per 48 min, “off-court”: -7.61
+/- per 48 min swing: 17.45 (I’d assume on/off is the same, given the proximity of the per 48 and 100 values here)


[rk2023]

https://champsorchumps.us/summary/nba

For those whom want to track +/- and on/off on a per-48 minute basis, highly recommend this site. That and BBR Game Logs to get individual player minutes makes the process of consolidating info much easier.

On that note, Justin - I am wondering if you have rotation/minute heat-maps on some of the 80s and 90s Bulls teams (apologies if they've already been posted in this thread).

[rk2023]

I ran through 1988 and 1991 Magic at a higher level (didn't adjust for pace or minutes yet, and filtered out the partial games).

1988 (48 Games):
8.38 +/- Per-Game with Magic, -2.36 +/- Per-Game without

1991 (47 Games):
5.38 +/- Per-Game with Magic, -1.83 +/- Per-Game without

The former is more in-line with the full 82 MOV/game Lakers results, FWIW.

Interesting. Can you convert to per 48?

Am working on it or at least plan to this evening; realized this formula was more of a 50-50 minute assumption. Plan to adjust for possessions afterwards, then later dive into opponent strength and so on for data visualization.

1988 (47 Games):
10.85 +/- Per-48 with Magic, -10.29 +/- Per-48 without

1991 (48 Games):
6.62 +/- Per-48 with Magic, -9.03 +/- Per-48 without

That should do it, as far as accuracy is concerned. Absolutely monster impact here

[Squared2020]

.

[homecourtloss]

Some David Robinson numbers

Thank you.

Squared— do you have any games of Bill Walton?

[Squared2020]

.

[OhayoKD]

Spoiler:

https://champsorchumps.us/summary/nba

For those whom want to track +/- and on/off on a per-48 minute basis, highly recommend this site. That and BBR Game Logs to get individual player minutes makes the process of consolidating info much easier.

On that note, Justin - I am wondering if you have rotation/minute heat-maps on some of the 80s and 90s Bulls teams (apologies if they've already been posted in this thread).

1985 Data:


1988 Data:


1991 Data:


1993 Data:


1996 Data:

So if I'm reading this right...

-> In 1988 most of his minutes are with Oakley
-> In 91 they ran platoon systems where all the starters were kept together (this was more common back then).
-> In 1993, Jordan stayed with the starters while Pippen was assigned to bench units.
-> 1992 is not listed but I think that is when they started staggering the two more — although it is possible that Jordan was the one playing with bench units that year given his on/off.

I am unclear whether these are specific to the playoffs.

[lessthanjake]

So if I'm reading this right...

-> In 1988 most of his minutes are with Oakley
-> In 91 they ran platoon systems where all the starters were kept together (this was more common back then).
-> In 1993, Jordan stayed with the starters while Pippen was assigned to bench units.
-> 1992 is not listed but I think that is when they started staggering the two more — although it is possible that Jordan was the one playing with bench units that year given his on/off.

I am unclear whether these are specific to the playoffs.

The y-axis for 1993 goes up to 1.0 games, so I think that’s just based on one game—which means obviously we can’t really say anything at all based on it.

These also aren’t playoffs—it’s a representation of Squared’s data, which, aside from including the 1985 playoffs, is all regular season games for Jordan.

[rk2023]

By chance do you have a slate of 1991 Chuck and Pat Ewing games?

[Squared2020]

.

[Squared2020]

.

[Squared2020]

.

[Squared2020]

.

[lessthanjake]

.

Are you able to provide charts of the Lakers’ lineups by seconds into the game for the 1984-1985 and 1987-1988 data? I’m particularly curious about the overlap rotationally between Magic and Kareem. If it would take a while to put together, then absolutely feel free to ignore. Either way, thanks so much for all this great data!

[Squared2020]

.

[colts18]

If I'm reading those charts right, it seems like Magic and Kareem had practically the same rotation.