RealGM

EDIT:Hah. Okay, I ran the numbers in order to try and determine On/Off. All this guy did was look at Chicago's team MOV per game, it lines up for every year.

He obviously made the error of then comparing those numbers with the actual on court +/- of other players.

From the Double Dribble Blog. Found the link on Jordan's BBR page.

Code: Select all

Season       Age    PER      TS%      RawPM

1986-87*     23     29.8     .56       +0.9
1987-88*     24     31.7     .60       +3.4
1988-89*     25     31.1     .61       +1.6
1989-90*     26     31.2     .61       +3.3
1990-91*     27     31.6     .61       +9.1
1991-92*     28     27.7     .58      +11.0
1992-93*     29     29.7     .56       +6.5
1995-96*     32     29.4     .58      +12.2
1996-97*     33     27.8     .57      +10.8
1997-98*     34     25.2     .53       +7.1

I commented asking where the data came from. It's encouraging that its out there.

The plus minus figures didn’t go back before 1986, so I couldn’t get Mike’s rookie year.

What does this mean? This person watched nearly 800 games (regular season + playoff) to determine this?

Dipper 13 wrote:

The plus minus figures didn’t go back before 1986, so I couldn’t get Mike’s rookie year.

What does this mean? This person watched nearly 800 games (regular season + playoff) to determine this?

The data must have been available somewhere. I find it hard to believe that this person even had access to all of those game, let alone was able to watch all of them.

SideshowBob wrote:

Dipper 13 wrote:

The plus minus figures didn’t go back before 1986, so I couldn’t get Mike’s rookie year.

What does this mean? This person watched nearly 800 games (regular season + playoff) to determine this?

The data must have been available somewhere. I find it hard to believe that this person even had access to all of those game, let alone was able to watch all of them.

Maybe there are text based play-by-play descriptions of games that he was able to parse through?

Just a guess.

ISB wrote:

SideshowBob wrote:Maybe there are text based play-by-play descriptions of games that he was able to parse through?

Just a guess.

Its possible, but even still, I wasn't aware that such data existed.

If that is on court +/-, then MJ was a negative plus/minus guy in 1988

Thanks for the post SB
obviously raw +/- doesnt tell much but its nice to see an attempt to getting impact stats from late 80s/early 90s. Maybe theres some hope for what else might surface in upcoming years

SideshowBob wrote:
Its possible, but even still, I wasn't aware that such data existed.

I was reading Jack McCallum's Dream Team book and he mentioned visiting the Lakers later in the 1988 season. Riley mentioned that Magic's plus/minus was lagging behind Bird's and MJ's that year. So someone must have had the data.

colts18 wrote:If that is on court +/-, then MJ was a negative plus/minus guy in 1988

If this data is accurate, we should be able to come up with off court estimates by estimating possessions based on Jordan's MP and the pace figures we already have available.

SideshowBob wrote:

ISB wrote:

SideshowBob wrote:Maybe there are text based play-by-play descriptions of games that he was able to parse through?

Just a guess.

Its possible, but even still, I wasn't aware that such data existed.

Same, I'm pretty confused. This means one of a few things I think:

1) PbP data exists for earlier seasons, but is either not publicly available or not digitalized (one of my colleagues from nbastats.net informed me that a poster "mtamada" on the APBR boards has mentioned that the Celtics used to record PbP back in the 80s).
2) He watched every available game of MJ (similar to Dipper 13's work with Barkley and some others, for which I haven't AND-1'd him enough ).
3) RawPM isn't actually +/-, but some proprietary metric of his.
4) Something more sinister is at play (fabrication of some sort).

I'm willing to give the guy the benefit of the doubt and say it''s (2), there's a decent chance it's (3), I'm hoping it's (1), and cautiously optimistic that it's not (4).

colts18 wrote:

SideshowBob wrote:
Its possible, but even still, I wasn't aware that such data existed.

I was reading Jack McCallum's Dream Team book and he mentioned visiting the Lakers later in the 1988 season. Riley mentioned that Magic's plus/minus was lagging behind Bird's and MJ's that year. So someone must have had the data.

Hmmm i have that book. Do u remember or know the page number?

SideshowBob wrote:

colts18 wrote:If that is on court +/-, then MJ was a negative plus/minus guy in 1988

If this data is accurate, we should be able to come up with off court estimates by estimating possessions based on Jordan's MP and the pace figures we already have available.

If his numbers are on court, not net +/- then I got this for 1988 assuming equal pace on and off court:

+3.4 on court
+4.2 off court
-0.8 Net +/-

That would go against what McCallum wrote in his book. So I assume that this net +/- though if it is, those numbers seem really low for MJ.

Pretty sure the plus-minus rating McCallum referred to is a completely different stat.

Sitting in his office in the Forum one day not long ago, Riley glanced at a computer printout that listed a "plus-minus rating" (a statistical analysis made by the Lakers' staff that covers everything from points scored to fouls committed) for every player in the NBA. The Chicago Bulls' Michael Jordan was on top with a .770 rating, Boston's Larry Bird was second at .700 and Johnson was in third place at .680.

"Last year Earvin was about .765, and I thought he'd play at that level this year, too," said Riley. "Last season he was a driven player. He was driven to win the MVP award and finally get his due. Because he's not a selfish player, it was hard for him, but he did it. He did it by constantly pushing himself to shoot, to penetrate, to take over a game. This year, for whatever reason, he hasn't done it. Maybe he's taken it upon himself to develop Byron. Maybe there's another reason. But I'd like to see him get that .765 every night."

http://sportsillustrated.cnn.com/vault/ ... /index.htm

GSP wrote:Hmmm i have that book. Do u remember or know the page number?

I just checked my copy of the book.

Page 31:

"I went to L.A. before the playoffs and Riley pulled out a computer printout showing the performances of Magic, Worthy, and Abdul-Jabbar had all gone down from the season before. Riley then put Magic in the context of Jordan and Bird, hauling out another stat sheet to reveal that Johnson's plus-minus rating was third in the league, below the metrics of the other two. Coaches and players hate when journalists make comparisons based on statistics, but they do exactly the same thing."

Hah. Okay, I ran the numbers in order to try and determine On/Off. All this guy did was look at Chicago's team MOV per game, it lines up for every year.

He obviously made the error of then comparing those numbers with the actual on court +/- of other players.

We should organize a group PbP/shot tracking project for key players prior to 97 a la Dipper13

NugzHeat3 wrote:Pretty sure the plus-minus rating McCallum referred to is a completely different stat.

Sitting in his office in the Forum one day not long ago, Riley glanced at a computer printout that listed a "plus-minus rating" (a statistical analysis made by the Lakers' staff that covers everything from points scored to fouls committed) for every player in the NBA. The Chicago Bulls' Michael Jordan was on top with a .770 rating, Boston's Larry Bird was second at .700 and Johnson was in third place at .680.

"Last year Earvin was about .765, and I thought he'd play at that level this year, too," said Riley. "Last season he was a driven player. He was driven to win the MVP award and finally get his due. Because he's not a selfish player, it was hard for him, but he did it. He did it by constantly pushing himself to shoot, to penetrate, to take over a game. This year, for whatever reason, he hasn't done it. Maybe he's taken it upon himself to develop Byron. Maybe there's another reason. But I'd like to see him get that .765 every night."

http://sportsillustrated.cnn.com/vault/ ... /index.htm

I would have loved to see the details behind this metric.

SideshowBob wrote:EDIT:Hah. Okay, I ran the numbers in order to try and determine On/Off. All this guy did was look at Chicago's team MOV per game, it lines up for every year.

He obviously made the error of then comparing those numbers with the actual on court +/- of other players. - See more at: viewtopic.php?f=64&t=1278468#sthash.VOmpliJ2.dpuf

From the Double Dribble Blog. Found the link on Jordan's BBR page.

Code: Select all

Season       Age    PER      TS%      RawPM

1986-87*     23     29.8     .56       +0.9
1987-88*     24     31.7     .60       +3.4
1988-89*     25     31.1     .61       +1.6
1989-90*     26     31.2     .61       +3.3
1990-91*     27     31.6     .61       +9.1
1991-92*     28     27.7     .58      +11.0
1992-93*     29     29.7     .56       +6.5
1995-96*     32     29.4     .58      +12.2
1996-97*     33     27.8     .57      +10.8
1997-98*     34     25.2     .53       +7.1

I commented asking where the data came from. It's encouraging that its out there.

It\s not real +/-, but some kind of estimations based on box score (that's why there's no stat for '85 - basketball-reference has full box scores only back to '86 season).

DavidStern wrote:

SideshowBob wrote:EDIT:Hah. Okay, I ran the numbers in order to try and determine On/Off. All this guy did was look at Chicago's team MOV per game, it lines up for every year.

He obviously made the error of then comparing those numbers with the actual on court +/- of other players. - See more at: viewtopic.php?f=64&t=1278468#sthash.VOmpliJ2.dpuf

From the Double Dribble Blog. Found the link on Jordan's BBR page.

Code: Select all

Season       Age    PER      TS%      RawPM

1986-87*     23     29.8     .56       +0.9
1987-88*     24     31.7     .60       +3.4
1988-89*     25     31.1     .61       +1.6
1989-90*     26     31.2     .61       +3.3
1990-91*     27     31.6     .61       +9.1
1991-92*     28     27.7     .58      +11.0
1992-93*     29     29.7     .56       +6.5
1995-96*     32     29.4     .58      +12.2
1996-97*     33     27.8     .57      +10.8
1997-98*     34     25.2     .53       +7.1

I commented asking where the data came from. It's encouraging that its out there.

It\s not real +/-, but some kind of estimations based on box score (that's why there's no stat for '85 - basketball-reference has full box scores only back to '86 season).

Nope, it's just team MOV. See the top of the opening post.

And BTW, it's impossible to do something like that manually (by watching all Jordan's games), because not all MJ's games are available. It's possible to do so with '96 season (games are relatively easily accessible), but rest of his career is out of reach for now.