APBRmetrics

The poster, Fpliii, was nice enough to go over an old Sixers media guide and take a picture of the section where it contained the raw plus/minus for every NBA player over 750 minutes.

http://tinypic.com/view.php?pic=15fnozr ... AKalPldU8w

http://tinypic.com/view.php?pic=8z21ig&s=8

Here is a spreadsheet of the on court, off court, and net plus/minus for 1994 players using B-R pace estimates

https://docs.google.com/spreadsheets/d/ ... edit#gid=0


Using that data I was able to come up with a regressed version of RAPM using 5 inputs: Minutes played, On court +/-, Off court +/-, PER, and WS/48 with 2014 Gotbuckets RAPM being the output variable. I came up with a formula with a R^2 of 0.62. Adding PER and WS/48 makes the r^2 value go from 0.59 to 0.62. More accurate, but by a small margin

Here are the results. The last column is that regressed RAPM adjusted to the variance of the 2014 RAPM

https://docs.google.com/spreadsheets/d/ ... 2018314684


-Most of the stars that year looked really good
-Robinson, Hakeem, and Malone stand out
-The Sonics were really good that year which makes the upset to the Nuggets even more shocking



EDIT: In post #4, there is 1995 and 1996 Plus/Minus data

That's awesome. Hopefully, one day, we'll have PBP for the 90s

I'm guessing McMillan is rated so high largely because of defense. He had a 5.8% STL% that year. Last seasons' leader (MP >1000) had a 3.8% STL%. A little surprised to see Shaq with "only" a +7 NET, which puts him at #29. Robinson, Malone, Olajuwon and Mutombo at the top are obviously no surprises

J.E. wrote:That's awesome. Hopefully, one day, we'll have PBP for the 90s

I'm guessing McMillan is rated so high largely because of defense. He had a 5.8% STL% that year. Last seasons' leader (MP >1000) had a 3.8% STL%. A little surprised to see Shaq with "only" a +7 NET, which puts him at #29. Robinson, Malone, Olajuwon and Mutombo at the top are obviously no surprises

PBP in some form has to go back to 87-88 at the least, since that's the first edition of the Sixers Media Guide with dunk data (unless they just counted dunks manually from watching games, as opposed to referring to the PBP). But yeah, would be amazing if the PBP for those three seasons is digitalized (as 96-97 through 99-00 recently were).

Anyhow though, I'm trying to get ahold of those two editions of Pollack's Statistical Yearbook. Raw plus/minus probably has limited utility, but it's still interesting data.

Fpliii was able to get the 95 and 96 season plus/minus results.

1995 Net Plus/Minus

https://docs.google.com/spreadsheets/d/ ... d=78057332

1995 Regressed RAPM results:

https://docs.google.com/spreadsheets/d/ ... edit?pli=1

1996 Regressed RAPM results

https://docs.google.com/spreadsheets/d/ ... edit?pli=1


Standouts:
-Robinson, Pippen, MJ, Stockton, Malone, Penny, Mookie, Zo
-Most of the 90's stars look good by the Net Plus/Minus metric

I ran a correlation between my 1995 Regressed RAPM and J.E.'s 1995 RAPM and I got a .718 R value.

fpliii wrote: ↑Tue Sep 02, 2014 5:54 am
PBP in some form has to go back to 87-88 at the least, since that's the first edition of the Sixers Media Guide with dunk data (unless they just counted dunks manually from watching games, as opposed to referring to the PBP). But yeah, would be amazing if the PBP for those three seasons is digitalized (as 96-97 through 99-00 recently were).

Anyhow though, I'm trying to get ahold of those two editions of Pollack's Statistical Yearbook. Raw plus/minus probably has limited utility, but it's still interesting data.

Yes. It does, it actually goes back a bit further. I have the Harvey Pollack's media guides all the way back to 1976-77 (1975-76 season) which has Plus/Minus for the 76ers team. In the 1977-78 media guide it has dunk data, after Dr. J's first year with the team. So Plus/Minus data might even go back any further, at this point I haven't seen the guides at a price point I'd like, but hope to collect at 44 media guides at some point.

pdevos wrote: ↑Tue May 25, 2021 10:22 pm

fpliii wrote: ↑Tue Sep 02, 2014 5:54 am
PBP in some form has to go back to 87-88 at the least, since that's the first edition of the Sixers Media Guide with dunk data (unless they just counted dunks manually from watching games, as opposed to referring to the PBP). But yeah, would be amazing if the PBP for those three seasons is digitalized (as 96-97 through 99-00 recently were).

Anyhow though, I'm trying to get ahold of those two editions of Pollack's Statistical Yearbook. Raw plus/minus probably has limited utility, but it's still interesting data.

Yes. It does, it actually goes back a bit further. I have the Harvey Pollack's media guides all the way back to 1976-77 (1975-76 season) which has Plus/Minus for the 76ers team. In the 1977-78 media guide it has dunk data, after Dr. J's first year with the team. So Plus/Minus data might even go back any further, at this point I haven't seen the guides at a price point I'd like, but hope to collect at 44 media guides at some point.

Haven’t been on here in a while, but just saw this. I actually collected the media guides from 96-97 back to the late 60s a while back. I can probably share the scans etc. There isn’t +/- prior to 75-76 however there is some team-level data on turnovers (and possibly offensive rebounds? I forget) league-wide beginning with 69-70.