PER Question.....

[THE HCP]

Can somebody find Brandon's PER for the 50+ game he had against PHX. I wonder how Dre's game tonight compares.

 

[Nikolokolus]

http://www.hoopdata.com/boxscore.aspx?id=300130006

I'm working on it, but in the meantime you can look at this.

 

[Nikolokolus]

http://www.basketball-reference.com...&c3val=&c4stat=&c4comp=gt&c4val=&order_by=pts

I can't seem to locate any single game PER stats, but off-hand I'd say it's pretty close to a dead heat. Miller's shooting percentage is ridiculous, but Brandon's number of free throws, assists, threes made and no turnovers might tip the balance in his favor. If you're feeling really bored you could always plug the numbers into Hollinger's forumla and see what turns up.

PER forumula

 

[Cake]

Can somebody find Brandon's PER for the 50+ game he had against PHX. I wonder how Dre's game tonight compares.

PER is normalized to compare a player's stats to league averages. So a single game PER isn't really possible. But game score is very close.

Brandon Roy's game score for his 52 point game was 44.3. Last night, Andre's game score was 41.2.

Game Score; the formula is PTS + 0.4 * FG - 0.7 * FGA - 0.4*(FTA - FT) + 0.7 * ORB + 0.3 * DRB + STL + 0.7 * AST + 0.7 * BLK - 0.4 * PF - TOV.

Game Score was created by John Hollinger to give a rough measure of a player's productivity for a single game. The scale is similar to that of points scored, i.e., 40 is an outstanding performance, 10 is an average performance, etc.

Damon's 54 point game: 40.1
Clyde's best game score was 45 on 48 points against the spurs in 1991.

 

[THE HCP]

Thanks man! Great stuff.

 

[Boob-No-More]

It's not PER, but Hollinger has another, simpler formula for grading single game performances. It's called Game Score (GmSc):

"Game Score; the formula is PTS + 0.4 * FG - 0.7 * FGA - 0.4*(FTA - FT) + 0.7 * ORB + 0.3 * DRB + STL + 0.7 * AST + 0.7 * BLK - 0.4 * PF - TOV. Game Score was created by John Hollinger to give a rough measure of a player's productivity for a single game. The scale is similar to that of points scored, i.e., 40 is an outstanding performance, 10 is an average performance, etc."

The game logs at basketball-reference.com only go back to 1986-87. So, they don't include Geoff Petrie's two 51-point games in 72-73. And, you can't search on GmSc. I searched for all games from 1986-87 where a Blazer cored at least 45 points. Since Clyde also tended to get a lot of rebounds, assists, steals and occasionally blocks, I also checked his game logs while he was a Blazer and found these several additional games where he scored less than 45, but had a GmSC of 40 or greater. This list only covers 1986-87 through present, and may not be exhaustive (it includes all players who scored 45 or more and all games where Clyde had a GmSc => 40.0), but here's what I found sorted by GmSc:

Clyde Drexler: PTS = 48 GmSC = 45.0
Brandon Roy: PTS = 52 GmSc = 44.3
Clyde Drexler: PTS = 41 GmSc = 44.2
Clyde Drexler: PTS = 48 GmSc = 42.5
Clyde Drexler: PTS = 42 GmSc = 41.8
Clyde Drexler: PTS = 43 GmSc = 41.3
Clyde Drexler: PTS = 41 GmSc = 41.3
Andre Miller: PTS = 52 GmSc = 41.2
Clyde Drexler: PTS = 50 GmSC = 40.5
Clyde Drexler: PTS = 42 GmSc = 40.2
Damon Stoudamire: PTS = 54 GmSc = 40.1
Darius Miles: PTS = 47 GmSc = 39.6
Kiki Vandeweghe: PTS = 48 GmSc = 35.0

BNM

 

[Boob-No-More]

PER is normalized to compare a player's stats to league averages. So a single game PER isn't really possible. But game score is very close.

Brandon Roy's game score for his 52 point game was 44.3. Last night, Andre's game score was 41.2.

Game Score; the formula is PTS + 0.4 * FG - 0.7 * FGA - 0.4*(FTA - FT) + 0.7 * ORB + 0.3 * DRB + STL + 0.7 * AST + 0.7 * BLK - 0.4 * PF - TOV.

Game Score was created by John Hollinger to give a rough measure of a player's productivity for a single game. The scale is similar to that of points scored, i.e., 40 is an outstanding performance, 10 is an average performance, etc.

Damon's 54 point game: 40.1
Clyde's best game score was 45 on 48 points against the spurs in 1991.

Beat me to it.

BNM

 

[THE HCP]

That Darius miles game was crazy! I think it was at Denver and he looked like LeBron in the 1st half!!!

 

[Boob-No-More]

Anybody have a link to boxscores for Geoff Petrie's two 51-point games from 1/20/73 and 3/16/73? If we can find the boxscores, we can calculate his Game Scores to see how they compare to the other top individual games in team history.

BNM

 

[Boob-No-More]

It would also be cool to see the Game Scores for some of Walton's best games. He wasn't the scorer Clyde was, but he usually shot a very high FG%, led the league in REB and BLK in '77, averaged 5.0 AST in '78 and averaged a STL per game as a Blazer. So, he could really fill out a boxscore in several areas. His PER of 24.8 in 1977-78 was better than Clyde's career best (24.1 in 1987-88) and leads me to believe he probably had a few games with a GmSC in the 40 range.

I can't find a complete boxscore (no BLK, STL, TOV or ORB) for Walton's 1973 NCAA championship game, but given that he had 44 points on 21-22 FG and also pulled down 13 REB, leads me to believe his GmSC would have around 50 for that game.

BNM

 

[jlprk]

Can somebody find Brandon's PER for the 50+ game he had against PHX. I wonder how Dre's game tonight compares.

How about this idea. Calculate efficiency (or effectiveness, whichever you want to call it) per minute for Roy in his Phoenix game divided by our whole team's efficiency per minute for that game. Do the same for Miller in his Dallas game. Use the NBA.com definition (it's easier but a little less accurate than the ESPN definition).

eff..min..player

049 043.82 Roy
141 240.00 team in Roy's game

049 042.23 Miller
118 265.00 team in Miller's game

Roy: (49 / 43.82) / (141 / 240) = 1.90334
Miller: (49 / 42.23) / (118 / 265) = 2.60579
2.60579 / 1.90334 = 1.369

Miller's PER would be around 37% higher than Roy's, which is a great difference. It's like one guy having a 40 PER in a game and the other having a 55. Big, big difference.

 

[jlprk]

http://www.hoopdata.com/boxscore.aspx?id=300130006

I'm working on it, but in the meantime you can look at this.

WHAT ARE THE CHANCES ???

Look at the ESPN box score URL !!!!

The 0130 must mean Jan. 30, and maybe one 0 means 2010, but what about the rest ???

http://sports.espn.go.com/nba/boxscore?gameId=300130006

This is beyond spooky !!!

 

[jlprk]

Roy: (49 / 43.82) / (141 / 240) = 1.90334
Miller: (49 / 42.23) / (118 / 265) = 2.60579

So Roy was 90% more efficient per minute than the whole Blazer team per minute (including himself). Miller was 161% more efficient. The % would be higher if I compared the player to the rest of the team, not including himself.

Like PER, my approximation of it 1) is per minute, 2) puts the player's stat in proportion to the whole team's stats, 3) doesn't include defense, and 4) doesn't include the scoring margin. I think PER factors in whether the player's team wins, but the Blazers won both games, so I didn't have to do so to compare the two games.

According to Basketball-Reference, Roy's Game Score in his game was 44.8, compared to Miller's of 41.2. But Game Score is very similar to efficiency (ESPN definition) and doesn't account for the player's proportion of overall team effort.

There is no database that shows the best Game Scores of all time, but ESPN maintains a ranking by efficiency. So you can find historical Game Score rankings 2 ways. 1) Boob searched the game logs of the best Blazer players of all time, looked for their Game Scores, and made his own list. 2) Assume that ESPN efficiency rankings will be in the same order as Game Score rankings, and use efficiency as a proxy. It's a reasonable assumption--look at the formulas and their difference:

Game Score
PTS + 0.4*FG - .7*FGA - .4*(FTA - FT) + .7*ORB + .3*DRB + STL + .7*AST + .7*BLK - .4*PF - TOV.

ESPN efficiency
PTS + REB + 1.4*AST + STL + 1.4*BLK -.7*TO + FGM + .5*FGM -.8*(FGA-FGM) + .25*FTM - .8*(FTA-FTM)

ESPN efficiency minus Game Score
REB - .7*ORB - .3*DRB + .7*AST + .7*BLK + .3*TO + .4*PF + 1.9*FGM - .1*FGA + .65*FTM - .4*FTA

 

[jlprk]

Here's an alternative method to find a single game PER. Just look at the player's season PER before and after the game and subtract to find the change from that one game. (You find it at either the Hollinger ESPN link, or Basketball-Reference.)

http://sportstwo.com/forums/showthread.php?p=2227729#post2227729

It's too late now to check Miller's pre-game PER, so I can't use the same method for his 52-point game.

 

[Nikolokolus]

Here's an alternative method to find a single game PER. Just look at the player's season PER before and after the game and subtract to find the change from that one game. (You find it at either the Hollinger ESPN link, or Basketball-Reference.)

http://sportstwo.com/forums/showthread.php?p=2227729#post2227729

It's too late now to check Miller's pre-game PER, so I can't use the same method for his 52-point game.

His PER was something like 16.3 before and it's a full point higher now.

 

[jlprk]

Then you use the same calculation BlazerCaravan used for Roy's Laker game in Post #12 of the last link I gave.

Before game: 21.4 PER x 1404 season minutes = 30045.6
After game: 22.0 PER x 1442 season minutes = 31724.0
31724.0 - 30045.6 = 1678.4 difference
1678.4 / 38 minutes played = 44.17 single-game PER

for Miller instead, in the Dallas game.

Before game: 16.3 PER x 1425 season minutes = 23227.5
After game: 17.3 PER x 1467 season minutes = 25379.1
25379.1 - 23227.5 = 2151.6 difference
2151.6 / 42 minutes played = 51 single-game PER