RealGM

https://www.sports-reference.com/blog/2020/04/nba-play-by-play-lineup-and-shooting-stats-added-back-to-1996-97/

Updated yesterday on Basketball Reference. PBP data was previously available on NBA.com but it is a lot easier to access now.
I have collated the on/off numbers of some of the more notable players below, with teammates grouped together to try to account for collinearity.

Key results:
• Karl Malone had the highest +/- numbers over this stretch with a 4 season average of +16.1 on/off in the regular season and +18.7 in the playoffs over 61 games. 97 Malone had the highest regular season on/off peak at +21.5 and his +27.8 postseason on/off is third only to 99 Robinson and 17 LeBron among players who reached the Finals. However, it is worth noting that rotations can heavily influence +/- numbers. If you look at the 97 and 99 Jazz as a team, every single bench player had a negative on/off while every starter was at a large positive. This happens when coaches don't stagger and often player all-starter or all-bench lineups which leads to a large disparity in +/- numbers within the team (see also 2006 Pistons, 2015 Clippers). Because his +/- is boosted by poor rotation management, Malone's adjusted +/- is not nearly as god-like as his raw +/- numbers would indicate.
• Jordan and Shaq with +18 on/off postseason averages, as expected from GOAT peak candidates.
• David Robinson had a +35.0 postseason on/off in the Spurs' 1999 championship run. Duncan was at -3.6 on/off.
• Kidd in 4 seasons before being traded to the Nets had on/offs of -3.5, +20.2, -0.9, -1.3. 1999 was a major outlier.
• Mourning, Mutombo, Payton, Garnett, Stockton, Blaylock, Duncan and Carter had several great +/- regular seasons.
• Over 3 years, Mookie Blaylock had an average on/off of +13.7 in the regular season and +26.3 in the playoffs (23 games). Both numbers were better than his teammate Mutombo's. However, Mookie only had a -3.4 on/off in 2000 after being traded whereas Mutombo maintained a +13.0 on/off. (Note: The 97 Hawks also had clear collinearity issues with poor rotation management.)

Regular season

Michael Jordan
97 RS: +13.3 On, +8.0 Net
98 RS: +8.4 On, +6.7 Net
Average: +10.8 On, +7.3 Net

Scottie Pippen
97 RS: +13.4 On, +8.1 Net
98 RS: +9.0 On, +3.3 Net
99 RS: +1.8 On, -5.2 Net
00 RS: +8.0 On, +0.5 Net
Average: +8.6, +2.2 Net

Karl Malone
97 RS: +11.6 On, +21.5 Net
98 RS: +8.9 On, +14.3 Net
99 RS: +10.4 On, +13.0 Net
00 RS: +8.2 On, +14.4 Net
Average: +9.7 On, +16.1 Net

John Stockton
97 RS: +8.7 On, +7.5 Net
98 RS: +12.3 On, +12.5 Net
99 RS: +11.6 On, +10.5 Net
00 RS: +10.2 On, +14.6 Net
Average: +10.4 On, +11.1 Net

Shaquille O'Neal
97 RS: +6.5 On, +4.8 Net
98 RS: +15.7 On, +14.9 Net
99 RS: +6.4 On, +10.0 Net
00 RS: +10.4 On, +10.9 Net
Average: +10.1 On, +10.4 Net

Kobe Bryant
98 RS: +7.4 On, -3.2 Net
99 RS: +2.9 On, -2.3 Net
00 RS: +9.9 On, +4.9 Net
Average: +7.1 On, +0.2 Net

David Robinson
98 RS: +7.8 On, +6.4 Net
99 RS: +11.4 On, +4.4 Net
00 RS: +8.8 On, +3.5 Net
Average: +9.0 On, +4.8 Net

Tim Duncan
98 RS: +8.0 On, +13.1 Net
99 RS: +10.6 On, +4.2 Net
00 RS: +9.9 On, +8.8 Net
Average: +9.3 On, +9.4 Net

Jason Kidd
97 RS: +2.7 On, +7.9 Net
98 RS: +4.7 On, -3.5 Net
99 RS: +3.4 On, +20.2 Net
00 RS: +5.5 On, -0.9 Net
Average: +4.2 On, +4.5 Net

Alonzo Mourning
97 RS: +10.7 On, +10.5 Net
98 RS: +13.0 On, +12.7 Net
99 RS: +12.4 On, +18.2 Net
00 RS: +7.1 On, +5.8 Net
Average: +10.4 On, +11.0 Net

Tim Hardaway
97 RS: +8.7 On, +11.7 Net
98 RS: +8.5 On, +9.7 Net
99 RS: +8.5 On, +4.0 Net
00 RS: +8.0 On, +4.6 Net
Average: +8.5 On, +8.4 Net

Grant Hill
97 RS: +7.3 On, +1.6 Net
98 RS: +4.1 On, +5.5 Net
99 RS: +3.8 On, +5.3 Net
00 RS: +2.2 On, +7.7 Net
Average: +4.5 On, +4.9 Net

Hakeem Olajuwon
97 RS: +7.0 On, +7.4 Net
98 RS: +2.1 On, +5.1 Net
99 RS: +2.1 On, -2.1 Net
Average: +4.3 On, +4.1 Net

Gary Payton
97 RS: +11.0 On, +8.3 Net
98 RS: +10.8 On, +9.6 Net
99 RS: +1.1 On, +8.5 Net
00 RS: +3.7 On, +14.7 Net
Average: +7.2 On, +10.5 Net

Allen Iverson
97 RS: -7.8 On, -0.9 Net
98 RS: -2.7 On, +1.4 Net
99 RS: +4.0 On, +7.9 Net
00 RS: +1.0 On, -0.3 Net
Average: -1.9 On, +1.5 Net

Reggie Miller
97 RS: +0.8 On, -2.8 Net
98 RS: +10.8 On, +11.5 Net
99 RS: +5.9 On, +5.2 Net
00 RS: +5.9 On, -1.7 Net
Average: +5.7 On, +2.7 Net

Dikembe Mutombo
97 RS: +8.5 On, +6.2 Net
98 RS: +9.7 On, +14.0 Net
99 RS: +7.0 On, +13.0 Net
00 RS: -1.5 On, +13.1 Net
Average: +5.8 On, +11.4 Net

Mookie Blaylock
97 RS: +10.7 On, +16.5 Net
98 RS: +9.7 On, +12.3 Net
99 RS: +6.8 On, +11.3 Net
00 RS: -10.4 On, -3.4 Net
Average: +4.6 On, +9.5 Net

Kevin Garnett
97 RS: +2.7 On, +15.0 Net
98 RS: +2.4 On, +6.8 Net
99 RS: +2.1 On, +8.6 Net
00 RS: +5.4 On, +11.6 Net
Average: +3.3 On, +10.7 Net

Vince Carter
99 RS: +1.4 On, +10.8 Net
00 RS: +4.0 On, +17.5 Net
Average: +3.5 On, +15.1 Net

Ray Allen
98 RS: -1.0 On, +4.2 Net
99 RS: +5.8 On, +6.5 Net
00 RS: +2.3 On, +8.5 Net
Average: +1.7 On, +6.3 Net

Chris Webber
97 RS: +5.0 On, +7.6 Net
98 RS: +0.1 On, +1.4 Net
99 RS: +1.1 On, +4.9 Net
00 RS: +1.2 On, -1.5 Net
Average: +1.9 On, +2.9 Net

Patrick Ewing
97 RS: +6.2 On, +7.4 Net
98 RS: +6.2 On, +4.5 Net
99 RS: +2.6 On, -0.1 Net
00 RS: +2.9 On, +1.8 Net
Average: +4.6 On, +4.1 Net

Postseason
(Note: Off court sample is extremely small for most of the runs so you will see some wild numbers here)

Michael Jordan
97 PO: +8.8 On, +23.6 Net
98 PO: +8.8 On, +13.1 Net
Average: +8.8 On, +18.2 Net

Scottie Pippen
97 PO: +8.3 On, +13.8 Net
98 PO: +6.7 On, -1.4 Net
99 PO: -2.7 On, -18.1 Net
00 PO: +8.6 On, +10.1 Net
Average: +7.0 On, +5.2 Net

Karl Malone
97 PO: +3.2 On, +27.8 Net
98 PO: +1.7 On, +13.2 Net
99 PO: +3.5 On, +15.4 Net
00 PO: -3.8 On, +24.4 Net
Average: +1.7 On, +18.7 Net

John Stockton
97 PO: -1.7 On, -2.2 Net
98 PO: -1.6 On, -2.3 Net
99 PO: +9.3 On, +24.4 Net
00 PO: +0.2 On, +24.9 Net
Average: +0.5 On, +7.0 Net

Shaquille O'Neal
97 PO: +10.4 On, +17.7 Net
98 PO: +7.7 On, +11.6 Net
99 PO: -3.5 On, +14.7 Net
00 PO: +4.7 On, +22.9 Net
Average: +5.1 On, +18.3 Net

Kobe Bryant
98 PO: -8.7 On, -22.2 Net
99 PO: -1.2 On, +29.7 Net
00 PO: +2.2 On, +0.1 Net
Average: -0.3 On, +3.3 Net

David Robinson
98 PO: +9.1 On, +18.4 Net
99 PO: +20.3 On, +35.0 Net
00 PO: +1.9 On, +25.2 Net
Average: +14.2 On, +28.4 Net

Tim Duncan
98 PO: +5.0 On, -3.1 Net
99 PO: +10.4 On, -3.6 Net
Average: +8.6 On, -3.4 Net

Jason Kidd
97 PO: -14.5 On, +6.8 Net
98 PO: -10.7 On, -13.7 Net
99 PO: -18.7 On, -31.3 Net
00 PO: -2.1 On, +4.4 Net
Average: -10.5 On, -5.3 Net

Alonzo Mourning
97 PO: +3.5 On, +14.7 Net
98 PO: +3.6 On, +16.6 Net
99 PO: -7.9 On, +1.0 Net
00 PO: +3.4 On, -1.3 Net
Average: +1.8 On, +8.4 Net

Tim Hardaway
97 PO: +1.9 On, +14.1 Net
98 PO: -4.5 On, -15.0 Net
99 PO: -14.8 On, -29.7 Net
00 PO: -7.7 On, -18.4 Net
Average: -2.9 On, -1.7 Net

Grant Hill
97 PO: -5.2 On, -9.9 Net
99 PO: -5.8 On, -10.0 Net
00 PO: -11.9 On, +7.9 Net
Average: -6.3 On, -7.7 Net

Hakeem Olajuwon
97 PO: +4.4 On, +7.9 Net
98 PO: -11.6 On, -21.4 Net
99 PO: -7.1 On, -17.6 Net
Average: -0.3 On, -1.3 Net

Gary Payton
97 PO: +7.2 On, +2.3 Net
98 PO: -1.0 On, +17.3 Net
00 PO: -4.2 On, -38.1 Net
Average: +2.1 On, +0.2 Net

Allen Iverson
99 PO: +2.7 On, +9.9 Net
00 PO: -5.4 On, -10.4 Net
Average: -1.8 On, -1.4 Net

Reggie Miller
98 PO: +4.8 On, +3.8 Net
99 PO: +3.0 On, -3.3 Net
00 PO: +6.1 On, +13.5 Net
Average: +5.0 On, +6.4 Net

Dikembe Mutombo
97 PO: -1.0 On, +4.9 Net
98 PO: -0.8 On, -11.5 Net
99 PO: -4.9 On, +2.5 Net
Average: -2.6 On, +1.5 Net

Mookie Blaylock
97 PO: +2.7 On, +56.0 Net
98 PO: +4.2 On, +5.7 Net
99 PO: -5.5 On, -1.4 Net
Average: -0.2 On, +26.3 Net

Kevin Garnett
97 PO: -16.0 On, -21.1 Net
98 PO: -6.8 On, +8.2 Net
99 PO: -11.7 On, -27.4 Net
00 PO: -2.4 On, +4.7 Net
Average: -8.6 On, -7.4 Net

Vince Carter
00 PO: -7.5 On, -12.2 Net

Ray Allen
99 PO: +3.8 On, +65.5 Net
00 PO: +9.7 On, +30.4 Net
Average: +7.4 On, +44.2 Net

Chris Webber
97 PO: +4.9 On, +23.3 Net
99 PO: -10.0 On, -22.6 Net
00 PO: -9.3 On, +0.1 Net
Average: -6.5 On, -3.8 Net

Patrick Ewing
97 PO: +6.1 On, +9.1 Net
98 PO: -18.4 On, -22.7 Net
99 PO: +10.8 On, +12.5 Net
00 PO: +1.0 On, +1.5 Net
Average: +3.0 On, +4.1 Net

Edit: Added Ewing.

Dikembe Mutombo stands out for me. I looked back at his career stats and forgot that he made 8 all-star games and won 4 defensive player of the year awards. He is a worthy selection to the hall of fame.

Mookie in Atlanta has always looked like an absolute stud by these types of numbers. One of histories missed stars.

I feel I remember somebody doing a deep dive on the '99 Spurs playoff #'s one time, but don't remember all the details, may need to do that myself.

eminence wrote:Mookie in Atlanta has always looked like an absolute stud by these types of numbers. One of histories missed stars.

I feel I remember somebody doing a deep dive on the '99 Spurs playoff #'s one time, but don't remember all the details, may need to do that myself.

I'm pretty sure you are referring to a post I made years back about Robinson.

colts18 wrote:David Robinson:
97: +7.62 on court, -9.37 off court, +16.99 plus/minus. 95.8 D rating on, 109.3 off, -13.5 Net D rating

The Spurs could have been decent if Robinson played that year instead of injured.

From 98-00, he averaged around a +7 plus/minus and -2.7 D rating impact.

But Robinson impressed the most in the playoffs. From 98-00:
on court: +12.66, 89.27 D rating
off court: -17.47, 102.52 D rating
Net: +30.12 plus/minus, -13.25 D rating impact

Duncan wasn't even close to Robinson in the playoffs plus/minus. Robinson was most impressive in the 99 finals run. On court, the Spurs were +19.5 with Robinson on court. An absurd number. They were -71 in the 215 minutes he was off the court or -24.76 per 100 possessions. His total net +/- was +35 and D rating impact was -16.25. Duncan was +9.5 on court compared to Robinson's +19.5. So the Spurs completely collapsed without Robinson in those playoffs, but played fine without Duncan.

The 99 season might be the all-time best defensive season ever for a player. According to B-R, the 99 Spurs finished with a -7.2 D rating relative to league average. The Spurs had a 91.0 D rating when Robinson was on the court (92.2 with Duncan). In the playoffs, they got even better. The Spurs had a 86.0 D rating when Robinson was playing on court. They allowed about 75 points per 48 minutes with Robinson on the court in the playoffs In the 2nd round, the Lakers had a 81 O rating when Robinson was playing. Shaq had one of his worst series ever going 23.8 PPG on .506 TS%. For all the talk of Robinson declining, he was actually really good in his final years. From 98-03, his plus/minus numbers were comparable to Duncan.

98-03 On Court Net +/-:
Drob: +9.04
Duncan: +8.35

Playoffs:
DRob: +8.12
Duncan: +6.65

Duncan had better Net +/-'s because of his minute totals, but Robinson wasn't too far off.

viewtopic.php?t=1234877

Hmm... Weak regular season #s from Robinson...

And no love for Ewing? The Ewing theory is basically dead now after seeing the +/- from the 99 playoffs.

Damn the Jazz were ass when Karl sat. This kills the narrative some have that Stockton was as impactful as Karl

Reggies numbers are underwhelming. Theyre still good but i thought hed be one of the kings of Apm and he clearly isnt

Maybe the MVP of Karl Malone in 1997 was deserved...

eminence wrote:I feel I remember somebody doing a deep dive on the '99 Spurs playoff #'s one time, but don't remember all the details, may need to do that myself.

colts had calculated some +/- stuff but without lineup stats, we were previously capped in the depth of analysis. With the new release, we can explore these types of Venn diagrams to break down the WOWY numbers:
BBR has also released the possession level ORtg and DRtg splits but the numbers don't seem reliable at the moment so I will wait to see if they get fixed. (Currently it has Robinson minus Duncan lineups playing at an 174 pace LOL).

E-Balla wrote:Hmm... Weak regular season #s from Robinson...

And no love for Ewing? The Ewing theory is basically dead now after seeing the +/- from the 99 playoffs.

To be fair, this is post-injury Robinson. From 94-96, he averaged a net +18.8 on/off in the regular season.

And my mistake, Ewing has been added to the list. I doubt anyone on here still believes in the Ewing theory anyway. It is pseudo-analytics peddled by people back when they didn't know anything about +/- or WOWY stats.

GSP wrote:Damn the Jazz were ass when Karl sat. This kills the narrative some have that Stockton was as impactful as Karl

Reggies numbers are underwhelming. Theyre still good but i thought hed be one of the kings of Apm and he clearly isnt

Stockton still has the edge in adjusted +/- over this period as per AScreamingComesAcrosstheCourt

97 NPI: 3.88 vs 4.42
98 NPI: 4.72 vs 4.63
99 NPI: 3.79 vs 2.92
00 NPI: 4.75 vs 1.96

98 RPI: 5.32 vs 5.31
99 RPI: 5.14 vs 4.58
00 RPI: 6.18 vs 4.12

And he is far ahead of Malone in ElGee's WOWY metrics. While it's nice to have these +/- numbers from BBR, I don't think they definitively point to Malone being the more impactful player on a per possession basis with the rotation issues on the team. APM was introduced precisely to solve the multicollinearity problem in raw +/- and it's telling that Malone drops off after these adjustments whereas other stars like Shaq, Jordan and Zo rise to the top of the league over him.

I agree on Reggie. I expected a little better from him especially since he had some decent sized playoff samples in 98 and 99.

Nice!

Can the stats section of NBA.com add play by play videos back to 96-97 as well?

The original small ball death lineup

MJ, Pippen, Harper, Kukoc, Rodman -> +55.7 in the playoffs! (Funny - at the time I always thought they should play that lineup more)

I really wish they had this data for the first threepeat also.

I thought the shot location note was interesting:

SHOOTING STATS: Shot charts and shooting statistics by distance have also been extended back to 1996-97 (previously these statistics went back to 2000-01 on the site). Before we point out where this data is, let us give a heavy disclaimer: SHOT LOCATION DATA FOR THE 1990s (PARTICULARLY 1996-97) IS SUSPECT. From 1996-97 through 1999-00 there are 194,239 field goal attempts for which the official play-by-play has no associated shot distance or location coordinates. For perspective, that is over a quarter of all field goals attempts for these seasons. Over 75% of these orphaned shots have been labeled as layups in the play-by-play accounts. The other most common descriptors for these shot attempts are dunk (over 15%) and tip shot (over 7%). Fewer than 2% are labeled as jump shots or hook shots. For the 1996-97 season, it is obvious that many missed jump shots were labelled as missed layups with no distance or coordinate data. So for 1996-97, in particular, and later 90s seasons to a lesser extent, field goal percentages at the rim and on layups are artificially deflated and field goal percentages on jump shots are artificially inflated. For instance, this missed Jordan turnaround fadeaway jumper is listed in official play-by-play as a missed layup. 1996-97 is the first season of the official digitized play-by-play, so it should come as no surprise that there are some issues with the data. It seems some teams may have been impacted by this more than others, with the 1996-97 Celtics (and David Wesley, in particular) having some very suspicious shooting splits by distance. We have decided to print these numbers, but we urge caution in utilizing them to prove any points.

GSP wrote:Damn the Jazz were ass when Karl sat. This kills the narrative some have that Stockton was as impactful as Karl

Reggies numbers are underwhelming. Theyre still good but i thought hed be one of the kings of Apm and he clearly isnt

They just peaked at diferent times. While Stockton was Jazz's best player from 88 to 92, Karl definitely was ahead by 94.

In 98 Stockton had already declined a lot. He still was very important for us tough.