Charles Barkley Synergy & Plus/Minus Statistics

[lorak]

What were his per game averages for that sample?

27.1 PPG, 56.8% FG

Really?... I reject these numbers. That's just too unbelievably good (jk)

In reality over that period of time Barkley averaged 25 ppg and 54.4 FG%. So not far away from Dipper's sample. And that sample has also playoff games and different amount of games from different years (so we shouldn't weight every season as equal to each other) - if we adjust for that, then maybe "real" Barkley's stats would be even more close to Dipper's sample.

[mysticbb]

Those D rating numbers off, that is clear as hell. We have data from Barkley's missed games, and not even in one of the samples we are getting such numbers. Seriously, either Dipper counted wrongly, or the bias in that sample is way too big towards his "better" games. If you believe, that those numbers are correct, you should place Barkley as one of the best defenders ever. Think about it, watch him play even in his best games, and then ask yourself whether you just saw a defender better than someone like Duncan, Garnett, etc. ;)

[Dipper 13]

^The On Court defensive rating isn't good at all (108.9), and in just the Sixers footage I believe the on court DRtg was between 112-113. But in both instances I remember the defense and the offense clearly was worse with him off the floor. There is no way he wasn't better on D than players like Chambers and Gilliam. The off court defensive numbers could certainly be skewed by numerous factors, I also found it almost impossible to believe. But for his defensive weaknesses and inconsistencies, he was certainly an underrated post defender in an era when it was very important.

[RebelWithACause]

Now I want to see Hakeem, MJ, Malone, and anyone else from that era ranked as high as Chuck (Ewing and The Admiral too).

Among a few other players, I am in the process of charting the Michael Jordan games. Any game footage I can find from 1990-1992 (his top 3 seasons in some order).

Great job here. When do you plan to finish Jordan?

[Dipper 13]

^ It could be a few weeks. Certainly before the new season begins.

[mysticbb]

^The On Court defensive rating isn't good at all (108.9), and in just the Sixers footage I believe the on court DRtg was between 112-113. But in both instances I remember the defense and the offense clearly was worse with him off the floor. There is no way he wasn't better on D than players like Chambers and Gilliam. The off court defensive numbers could certainly be skewed by numerous factors, I also found it almost impossible to believe. But for his defensive weaknesses and inconsistencies, he was certainly an underrated post defender in an era when it was very important.

We have a 15 game sample in 1991 without Barkley and 67 games with him. At the top of his play, his highest PER, no down season at all. Here is the result:

With him: 108.4 Ortg vs. 108.4 DRtg
Without him: 106.4 ORtg vs. 107.2 DRtg

There is really NOTHING, which makes it seem as if Barkley was some out of control defensive impact player like your numbers suggest. ;)

[Dipper 13]

We have a 15 game sample in 1991 without Barkley and 67 games with him. At the top of his play, his highest PER, no down season at all. Here is the result:

With him: 108.4 Ortg vs. 108.4 DRtg
Without him: 106.4 ORtg vs. 107.2 DRtg

There is really NOTHING, which makes it seem as if Barkley was some out of control defensive impact player like your numbers suggest.

Certainly he was not a great defender, he never approached an All NBA Defensive Team (at a time before they gave out flagrantly undeserved selections), but he was no liability. I have seen posters here declare him the absolute worst defensive player of his era. Both stretches when Barkley missed games that season, he was replaced in the starting lineup by the (streaky) 6th man Ron Anderson, who played excellent. The 1991 season was subpar for the Sixers as a team and Barkley himself in terms of health and durability. Skill wise he was the same, but it was a very turbulent season for the team who had 18 different players on the roster that season. The 1991 team had numerous problems, first being their starting PG going down early in the season with a torn ACL. They had to rely on 36 year old Ricky Green & Andre Turner to play the point. Those teams never had much depth to begin with, and the 6th man Ron Anderson was a streaky shooter. But the killer was the Gminski for Gilliam trade which didn't make sense to me, seeing as they traded a center who could provide some spacing with his outside shot vs. yet another inside player (Gilliam) to play alongside Mahorn and Barkley. Especially considering Hersey Hawkins was the only quality 3 point shooter on the roster. How about the idea of players doing things they don't normally do to compensate for such a loss? Or a team taking the game for granted (Bulls) knowing Barkley is out.





April 02, 1991

Charles Barkley and his sprained left knee will indeed be unavailable to the 76ers for a minimum of two weeks, but put heavy emphasis on the word minimum.

"I think we're going to be doing things different more than doing things better," Anderson said. "We're going to have to be trapping more and pushing the pace of the game, because we can't afford to play half-court basketball now. Fifty percent of our offense, maybe more, when we get into the half- court, is Charles Barkley. That's going to be missing. We have to trap effectively, be aggressive, go for every rebound. I would say we can do it. Whether we will or not, we'll see."





April 10, 1991

So how did the Sixers manage to win in Chicago? Charles Barkley offered this theory:

"They had a letdown when they had nothing to play for, because they won the (Central) division title the night before, when Detroit lost. And their best player (Michael Jordan) went off (golfing) at 7 in the morning (before a 2:30 p.m. game). I'd have gotten killed for (doing something like) that.

"They took the game for granted. I don't like when people are disrespectful to our team. They (the Bulls) felt they were going to win, that all they had to do was show up. Michael (despite scoring 41 points) did not play like Michael. (But) if Chicago came here without Michael, I'd be relaxed, too. It's human nature."

A source close to the Bulls said it is not unusual for Jordan to use golf as a form of relaxation before a game.





April 02, 1991

Milwaukee is an unlucky penny that keeps reappearing on the Sixers' schedule. Johnny Dawkins, the Sixers' starting point guard, was lost for the season to a knee injury in a November game against the Bucks, and Barkley suffered a minor ankle fracture in January against them.

The ankle injury cost Barkley seven games. Without Barkley for that stretch, the Sixers went 2-5 but played only two bad games. Their offense was all right in his absence, but their defense and rebounding suffered.

To be in better shape this time, the Sixers will have to get a stepped-up performance from power forward Armon Gilliam, whose recent rebounding has been less than wonderful. In the last 11 games, Gilliam has averaged only 4.7 rebounds.

"I need to be more involved," he said. "I need to be a more productive basketball player. Being without Charles is not really going to change my role, but I have to be more consistent.

"I can't do it all by myself. It will take two or three guys to fill the void Charles leaves."

Barkley was averaging 28.4 points and 10.3 rebounds before being injured Sunday. He was the Sixers' only effective offensive rebounder, and his absence will put a much higher premium on the team's getting and making good shots.

[mysticbb]

How about the idea of players doing things they don't normally do to compensate for such a loss? Or a team taking the game for granted (Bulls) knowing Barkley is out.

That might be it, but usually such effect is small and only during a couple of games, not 15, especially when they played the 76ers multiple times without Barkley like the Knicks, Pacers or Hornets.

We have a biased sample here, because the strength of schedule is not accounted for, and Barkley missed more home than road games. If we adjust those numbers for the opponents offensive and defensive strength as well as HCA, we get:

Without: -3.5 off per 100, +1.2 def
With: +1.0 off, -0.7 def

In order to achieve such difference, we would need a player with +5.8 off and -2.5 def in comparison to his replacement player.
So, the offensive impact was most certainly underrated by the previous presented numbers, but there is really nothing to suggest that Barkley had a big impact defensively. We can also look at the Suns, when he joined them. They got better offensively, but slightly worse defensively. Exactly what we expect from Barkley, an awesome offensive player, but below average defensively.

I agree, he wasn't the worst defensive player of his era, but among the superstars, he was rather at the bottom.

His performance level in the 67 games he played does not suggest a huge effect of his health issues.

[Dipper 13]

^Perhaps if the RAPM statistics were available from that time, the Off court defensive numbers would look a bit more as expected (certainly not 126 DRtg). I did not keep track of every lineup on the floor. My main reason for posting these On/Off Numbers is that a lot of posters here take a great deal of joy and pleasure in using statistics against the older players (pre-2000), when they just did not exist at the time. If the retired player is not a designated favorite of the forum, then very seldom will he get the benefit of the doubt in any comparison or discussion.

[lorak]

To add some information about Barkley's impact:

with/without

Code: Select all

            with         without         differential         
sezon   ortg   drtg      ortg   drtg      ortg   drtg   net   G without
1987   109,4   109,1      105,4   107,9      4,0   1,2   2,8   14
1991   108,8   108,8      106,4   107,6      2,4   1,2   1,2   15
1994   113,1   109,1      106,4   110,5      6,7   -1,4   8,1   17
1995   115,8   111,7      110,8   106,9      5,0   4,8   0,2   14
1996   111,2   110,1      108,7   112,8      2,5   -2,7   5,2   11
1997   110,7   104,6      107,2   104,6      3,5   0,0   3,5   29
1998   109,8   109,4      101,1   107,7      8,7   1,7   7,0   14
1999   107,1   103,6      101,8   103,8      5,3   -0,2   5,5   8




on/off (based on stats.nba.com data):

Code: Select all

        on the floor       off the floor      difference      
year   ortg   drtg      ortg   drtg      ortg   drtg   net
1997   108,2   100,7      103,3   102,1      4,9   -1,4   6,3
1998   109,7   107,3      97,7   103,5      12,0   3,8   8,1
1999   107,1   102,4      94,9   96,3      12,2   6,1   6,1
2000   99,7   102,0      102,0   103,3      -2,3   -1,3   -1,0
PLAYOFFS                           
1997   110,3   108,3      106,5   108,3      3,8   0,0   3,8
1998   100,9   103,6      89,6   99,0      11,3   4,6   6,8
1999   102,3   107,0      108,5   95,2      -6,2   11,8   -18,0



On/off ortg in 1998 and 1999 is at offensive GOAT level.

[Dipper 13]

To add some information about Barkley's impact:

with/without

Code: Select all

            with         without         differential         
sezon   ortg   drtg      ortg   drtg      ortg   drtg   net   G without
1987   109,4   109,1      105,4   107,9      4,0   1,2   2,8   14
1991   108,8   108,8      106,4   107,6      2,4   1,2   1,2   15
1994   113,1   109,1      106,4   110,5      6,7   -1,4   8,1   17
1995   115,8   111,7      110,8   106,9      5,0   4,8   0,2   14
1996   111,2   110,1      108,7   112,8      2,5   -2,7   5,2   11
1997   110,7   104,6      107,2   104,6      3,5   0,0   3,5   29
1998   109,8   109,4      101,1   107,7      8,7   1,7   7,0   14
1999   107,1   103,6      101,8   103,8      5,3   -0,2   5,5   8




on/off (based on stats.nba.com data):

Code: Select all

        on the floor       off the floor      difference      
year   ortg   drtg      ortg   drtg      ortg   drtg   net
1997   108,2   100,7      103,3   102,1      4,9   -1,4   6,3
1998   109,7   107,3      97,7   103,5      12,0   3,8   8,1
1999   107,1   102,4      94,9   96,3      12,2   6,1   6,1
2000   99,7   102,0      102,0   103,3      -2,3   -1,3   -1,0
PLAYOFFS                           
1997   110,3   108,3      106,5   108,3      3,8   0,0   3,8
1998   100,9   103,6      89,6   99,0      11,3   4,6   6,8
1999   102,3   107,0      108,5   95,2      -6,2   11,8   -18,0



On/off ortg in 1998 and 1999 is at offensive GOAT level.

I wouldn't be surprised if the NBA has all this data going back at least to 1990. If they're not going to release the old game footage from the 50's and 60's, then they should release these statistics going back further than 1997.

[mysticbb]

I wouldn't be surprised if the NBA has all this data going back at least to 1990. If they're not going to release the old game footage from the 50's and 60's, then they should release these statistics going back further than 1997.

I don't think it is available in a proper format to release the data, otherwise they would do it. The old data from 1997 to 2000 is also not in the best format (players are only named by the surname in the pbp, which makes it difficult to make a proper analysis for example). They may also be a few missing sequences. The data on NBA.com is also not consistent with what the stats community considers a possession, so there are few questionmarks here.

Anyway, regarding Barkley: As you can see, we see a rather consistent trend for him to be a net negative player on the defensive end, while being an offensive juggernaut. That is in agreement at least with my "eye test", where I see his faults on the defensive end causing enough trouble to lower the chances for a team on the defensive end to be great. His offensive impact shouldn't be in question, but his value as a team building cornerstone can be, as multiple examples of him causing trouble suggest.
Anyway, I have him as a +7.5 player on offense at his peak, while being -1 on defense for an overall +6.5 peak. In average I have him as a +5.5 player for his prime (about +7 on offense and -1.5 on defense). That would be better than Karl Malone for example, but Malone beats him in terms of longevity. Only Duncan, Garnett and Nowitzki are considered better PF in my list, while the difference between Barkley and Nowitzki is obviously the smallest among those.

[acrossthecourt]

I wouldn't be surprised if the NBA has all this data going back at least to 1990. If they're not going to release the old game footage from the 50's and 60's, then they should release these statistics going back further than 1997.

I don't think it is available in a proper format to release the data, otherwise they would do it. The old data from 1997 to 2000 is also not in the best format (players are only named by the surname in the pbp, which makes it difficult to make a proper analysis for example). They may also be a few missing sequences. The data on NBA.com is also not consistent with what the stats community considers a possession, so there are few questionmarks here.

Anyway, regarding Barkley: As you can see, we see a rather consistent trend for him to be a net negative player on the defensive end, while being an offensive juggernaut. That is in agreement at least with my "eye test", where I see his faults on the defensive end causing enough trouble to lower the chances for a team on the defensive end to be great. His offensive impact shouldn't be in question, but his value as a team building cornerstone can be, as multiple examples of him causing trouble suggest.
Anyway, I have him as a +7.5 player on offense at his peak, while being -1 on defense for an overall +6.5 peak. In average I have him as a +5.5 player for his prime (about +7 on offense and -1.5 on defense). That would be better than Karl Malone for example, but Malone beats him in terms of longevity. Only Duncan, Garnett and Nowitzki are considered better PF in my list, while the difference between Barkley and Nowitzki is obviously the smallest among those.

You mean how they calculate possessions that they use for instance in ORtg?

[mysticbb]

You mean how they calculate possessions that they use for instance in ORtg?

Well, I don't know whether they count or calculate possessions in their new database, but for both they are either counting or calculating more possessions than someone like me would count (or calculate). So, yes, the ORtg for instance are off on nba.com.
For Net values that obviously doesn't matter as much.

[lorak]

I talked with them (NBAstats) and they calculate possessions using simple formula (FGA + (0.44 x FTA) + TO – OReb). But what they do wrong way is how they calculate ortg/drtg. Normally to count ortg/drtg in one game we sum team's A possessions with team's B possessions, divide that by 2 and then calculate ortg/drtg. NBAstats skips "divide by 2" part and that causes strange results. For example here's how it looks for 2013 finals G1:

according to b-r
SAS pace 85
SAS ortg 108,2
SAS drtg 103,5
SAS +4,7 diff

according to NBAStats
SAS pace 87,7
SAS ortg 102,3
SAS drg 102,9
SAS -0,6 diff

[Dipper 13]

where I see his faults on the defensive end causing enough trouble to lower the chances for a team on the defensive end to be great.

I think he would still be a good low post post defender 1 on 1, except against the taller players like Gasol and Nowitzki, who don't rely on the low post positioning nearly as much as some others and could just shoot over him. His main problems were based on him being an interior player acting as the last line of defense. It seems the very advantage that gave him an edge over the slower footed big men offensively, was a disadvantage on defense. He could make some incredible weakside blocks every now and then, but for the most part he was very late in his rotations, if he even showed up at all. During his draft day interview, he even said Knight cut him from the Olympic team due to his subpar defense.

http://www.youtube.com/watch?v=juGvOPI97yk&t=2m28s


But in this era, with the new defensive rules and "zone" defenses, it would be easier for him to recover defensively and shade, though his physicality may work against him and he would still have trouble with the P&R. He would most likely be guarding the SF position today.



Dallas Morning News - Mar 1, 1990

Philadelphia's interior defense has improved dramatically, and Mahorn's presence has allowed Barkley to guard the small forward. "Charles has become more of a leader each year I've been here,' said Lynam, who has been in Philadelphia for the last three years. "He's putting more of a conscious effort into being involved at the defensive end of the court on a consistent basis. He's more involved with the other players during the game. "We need an involved, active Charles. That has a positive effect." Barkley says he likes this Sixers team. He joins Mahorn and Gminski to give them one of the most physical front lines around. Hawkins and Johnny Dawkins, obtained from San Antonio for veteran Maurice Cheeks in the off-season, complement each other in the backcourt.

[E-Balla]

I talked with them (NBAstats) and they calculate possessions using simple formula (FGA + (0.44 x FTA) + TO – OReb). But what they do wrong way is how they calculate ortg/drtg. Normally to count ortg/drtg in one game we sum team's A possessions with team's B possessions, divide that by 2 and then calculate ortg/drtg. NBAstats skips "divide by 2" part and that causes strange results. For example here's how it looks for 2013 finals G1:

according to b-r
SAS pace 85
SAS ortg 108,2
SAS drtg 103,5
SAS +4,7 diff

according to NBAStats
SAS pace 87,7
SAS ortg 102,3
SAS drg 102,9
SAS -0,6 diff

So basically they don't take having more possessions into account which is technically fine but then they'd need to list possessions and possessions against everytime they mentioned ORTG/DRTG.