Michael Jordan -- Complete Finals Plus Minus

[Djoker]

I've recently been watching a lot of Bulls playoff games and tracking plus minus for games before 1996-97.

Here is the complete game-by-game Michael Jordan plus-minus data for the NBA Finals:



I also have the points scored for and against and can post ORtg and DRtg later.

[Heej]

Is 35 games as a sample considered good enough to draw meaningful conclusions from? Cuz the variation is crazy and the ultimate net is somewhat hilarious

[Colbinii]

Is 35 games as a sample considered good enough to draw meaningful conclusions from? Cuz the variation is crazy and the ultimate net is somewhat hilarious

1500 Minutes isn't great but not horrible. 200 Minutes is quite the insignificant sample.

[zimpy27]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

[70sFan]

Are these estimated per 100 poss or per 48 min?

[jalengreen]

Are these estimated per 100 poss or per 48 min?

100 poss. 100*(Raw ON or OFF/Possessions) is the calculation, where Possessions=[(Pace/48)*Minutes ON or OFF]}.

[Djoker]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

It's not really bad IMO. I would imagine that in the Finals since you're playing the best opponents ON Net Rating usually isn't that high... And thus ON-OFF also won't be high.

Are these estimated per 100 poss or per 48 min?

per 100

[lessthanjake]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

I think +6.2 ON is actually extremely good for the finals. Let’s remember that the Bulls were playing great teams and may have won every series but they typically went to 6 games, so the other teams were still winning games.

Tim Duncan is the only person I can think of in the last 50 years that played in a lot of Finals and might plausibly have a similar ON rating in the Finals. Some quick math using basketball-reference +/- numbers tells me Duncan’s ON rating per 48 minutes in the Finals was a little below these numbers for Jordan (+5.60), but per-100-possession numbers are a little higher than per-48-minutes, so Duncan is surely very similar to Jordan in per-100-possession terms (perhaps very slightly ahead). And, perhaps not coincidentally, their teams had virtually identical winning percentages in the Finals (68% and 69%). I’ll note that even if going back further, given the Celtics’ average margin of victory in the games Bill Russell played, the high minutes he played, and the super high pace of that era, it’s essentially certain Bill Russell had a lower ON rating in the Finals than Jordan. In Finals games Russell played, the Celtics outscored their opponents by 338 points. Russell played 3181 minutes in those games and missed 218 minutes. And the Celtics’ pace in those years averaged 125.8. Even if we assume only 120 pace in the Finals, the Celtics would’ve needed to be around -29.0 per 100 possessions with Russell off the court for Russell to have had the ON rating per 100 possessions that Jordan had. Obviously, that’s extremely unlikely. Of course, this is not exact math (since pace might’ve been a little different with Russell on and off the court), but the basic message is the same: It is virtually certain that Bill Russell’s ON rating in the Finals was lower than Jordan’s.

Furthermore, let’s remember that Jordan was doing this against some really great teams. Jordan’s Finals opponents averaged 6.84 SRS in those seasons, and had major stars on their teams that were typically sharing the court in Jordan’s minutes. For reference, Duncan’s Finals opponents averaged only 3.95 SRS, while Russell’s Finals opponents averaged just 2.17 SRS. I’d say putting up a similar ON rating in the Finals to this board’s #5 all-time player and almost certainly a higher ON rating in the Finals than this board’s #4 all-time player, while playing against Finals opponents that were substantially stronger is extremely impressive. Not surprising, since we know his teams won every Finals he was in, but it’s definitely GOAT-like. The bottom line is that, when adjusted for opponent SRS (i.e. raw ON rating + Opponent SRS), I think Jordan’s Finals ON rating is almost certainly the highest of anyone in history that played in a bunch of Finals, and it’s likely not even particularly close. These numbers are *extremely* impressive, but are just unsurprising since we already know about Jordan’s incredible Finals dominance.

Personally, I find the OFF rating here a bit less interesting, just because the minutes sample is inherently so low. Obviously, the Bulls did very well with Jordan off the court in a couple series and very badly in others, but the minutes sample in individual series is tiny and even the total minutes across all Finals is too low to draw a real conclusion from IMO.

[Colbinii]

I wonder who has the largest On/Off in NBA Finals History?

Jordan is at +6.8
LeBron is at +15.9
Curry is at +7.4

Duncan's could be really high [don't have time to do his]. I imagine someone like Kawhi's to be really strong as well as Kobe's.

[Colbinii]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

I think +6.2 ON is actually extremely good for the finals. Let’s remember that the Bulls were playing great teams and may have won every series but they typically went to 6 games, so the other teams were still winning games.

Curry is at +5.3 ON.

Hard to imagine players who were on the losing end would have an ON remotely close to Curry/Jordan, since their teams were so dominant in the Finals.

[lessthanjake]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

I think +6.2 ON is actually extremely good for the finals. Let’s remember that the Bulls were playing great teams and may have won every series but they typically went to 6 games, so the other teams were still winning games.

Curry is at +5.3 ON.

Hard to imagine players who were on the losing end would have an ON remotely close to Curry/Jordan, since their teams were so dominant in the Finals.

A lot of the reason their teams were so dominant in the Finals is that they were so good with those players on the court. How dominant the teams were is not some independent variable here. Other players’ teams did not do as well in the Finals when they were on the court and that is a huge reason that their teams did not do as well overall.

[Colbinii]

I think +6.2 ON is actually extremely good for the finals. Let’s remember that the Bulls were playing great teams and may have won every series but they typically went to 6 games, so the other teams were still winning games.

Curry is at +5.3 ON.

Hard to imagine players who were on the losing end would have an ON remotely close to Curry/Jordan, since their teams were so dominant in the Finals.

A lot of the reason their teams were so dominant in the Finals is that they were so good with those players on the court.

Oh of course. It is why Jordan is unanimously a Top 3 player of all-time and Curry is in discussions with Magic as a Top 10 player and the greatest Point Guard ever.

How dominant the teams were is not some independent variable here.

Correct, the only thing I would say is it is easier to be +5-+6 On-Court when your team is neutral to -2 without you. That's fairly obvious though.

Other players’ teams did not do as well in the Finals when they were on the court and that is a huge reason that their teams did not do as well overall.

Yup, but at the end of the day basketball is played 5-on-5 and teammates do impact on-court results, just not as much as these ATG players.

[zimpy27]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

It's not really bad IMO. I would imagine that in the Finals since you're playing the best opponents ON Net Rating usually isn't that high... And thus ON-OFF also won't be high.

Are these estimated per 100 poss or per 48 min?

per 100

Not saying bad. Just a lot worse than I thought

[lessthanjake]

Curry is at +5.3 ON.

Hard to imagine players who were on the losing end would have an ON remotely close to Curry/Jordan, since their teams were so dominant in the Finals.

A lot of the reason their teams were so dominant in the Finals is that they were so good with those players on the court.

Oh of course. It is why Jordan is unanimously a Top 3 player of all-time and Curry is in discussions with Magic as a Top 10 player and the greatest Point Guard ever.

How dominant the teams were is not some independent variable here.

Correct, the only thing I would say is it is easier to be +5-+6 On-Court when your team is neutral to -2 without you. That's fairly obvious though.

Other players’ teams did not do as well in the Finals when they were on the court and that is a huge reason that their teams did not do as well overall.

Yup, but at the end of the day basketball is played 5-on-5 and teammates do impact on-court results, just not as much as these ATG players.

I agree with everything you said here

[ty 4191]

I've recently been watching a lot of Bulls playoff games and tracking plus minus for games before 1996-97.

Here is the complete game-by-game Michael Jordan plus-minus data for the NBA Finals:



I also have the points scored for and against and can post ORtg and DRtg later.

I'd be very interested to see the ORtg and DRtg for Jordan in the Finals.

Great work!! Curious...how long did this take you?

1. How does Jordan's on/off per 100 possessions compare to everyone else in the Finals (since 1996-1997)?

2. How about his ORtg and DRtg?

Thanks, everyone.

[homecourtloss]

I wonder who has the largest On/Off in NBA Finals History?

Jordan is at +6.8
LeBron is at +15.9
Curry is at +7.4

Duncan's could be really high [don't have time to do his]. I imagine someone like Kawhi's to be really strong as well as Kobe's.

I don’t know about Duncan – I think they were at least three or four finals in which the Spurs were better with Duncan off court or at least neutral.

As for Kawhi, the Raptors were +23.0 per 100 possessions with Kawhi off the court in 2019. These are the benefits of having all plus players in your rotation. Kawhi was +3.6 on-off in 2013, so he won’t be up on this list.

KD had a wild on - off in 2018 with a small 32 minute off sample but the Warriors and the thunder were both positive with him off court and 2017 and 2012, respectively

How dominant the teams were is not some independent variable here.

Correct, the only thing I would say is it is easier to be +5-+6 On-Court when your team is neutral to -2 without you. That's fairly obvious though.

Agree. +6.2 it’s really solid, though, even if the competition really isn’t that amazing.

[zimpy27]

Pippen comparison would be interesting

[Colbinii]

Pippen comparison would be interesting

Correct. And rotations in previous eras were typically closer to Starters/Bench than now-a-days.

[OhayoKD]

That's a lot worse than I imagined considering Bulls won each of these finals series and he played so many minutes..

The few minutes part doesn't neccesarily work in a players favor, teams can hold up better for a few minutes than they would for extended stretches depending on the circumstances.

Jordan tends to look better when we look at a few minutes than full-games(possibly a result of platooning based on the rotation charts) but without double checking how the bulls specifically performed in final games, the reverse seems to happen here for some reason.

Luckily we have an unusual abundance of larger samples with Jordan so we don't need to overindex on on/off here.

These numbers are bad relative to where he's typically argued but I wouldn't put too much stock in them

[Djoker]

...

Duncan's ON is quite high. He's +157 for his Finals career compared to +169 for MJ and his minutes average is much lower than MJ's so when we do the * 48/MP * 100/Pace he should come out pretty far ahead. Duncan for his career is probably around +7 to +8 ON Net in the Finals. If we exclude his non-superstar series which are 2013 and 2014, then he's behind Jordan.

Pippen comparison would be interesting

Correct. And rotations in previous eras were typically closer to Starters/Bench than now-a-days.

Actually not really.

A lot of Jordan's minutes on the bench, Pippen was on the floor and vice versa. In fact, this was the case in just about all non-blowout situations. In the 1996 Finals for instance, Pippen was on the court for 32 out of the 36 minutes MJ sat.

I'd be very interested to see the ORtg and DRtg for Jordan in the Finals.

Great work!! Curious...how long did this take you?

1. How does Jordan's on/off per 100 possessions compare to everyone else in the Finals (since 1996-1997)?

2. How about his ORtg and DRtg?

Thanks, everyone.

It's not so bad though tracking plus-minus. Just note the time and score when a player is subbed in and out. Adds a couple of minutes to watching a game but nothing unmanageable. I also make a few notes on defense and other stuff when I watch these games.

Here is an updated figure with the ORtg and DRtg.



And team and opponent points for each series with Jordan ON and Jordan OFF.