Diesel and the EFF Stat....

[Diesel]

First by their own definition:

Quote:

---

How do many NBA coaches quickly evaluate a player's game performance? They check his efficiency.

NBA.com evaluates all players based on the efficiency formula: ((Points + Rebounds + Assists + Steals + Blocks) - ((Field Goals Att. - Field Goals Made) + (Free Throws Att. - Free Throws Made) + Turnovers)).

For example, compare the following stat lines:

MIN FGM-A FTM-A REB AST STL BS TO PTS

Player A 43 5-22 7-9 8 6 3 0 4 17

Player B 29 5-8 3-4 4 7 0 0 2 15

Player A had a better game, right? Not exactly. Player B, who shot 5-8 from the field and committed two turnovers, registered a +20 efficiency total while Player A, who shot 5-22 from the field and committed four turnovers, posted a +11.

OK, so that's the way it's done. Me myself, believe that rebounds, Assists, Steals, and Blocks should be weighted slightly higher than points... however.. this is it.

At Home:

JSmoove = 22.23

JJ = 21.2

Al = 17.5

Marvin = 16.6 (21.5/48)

Chillz = 15.8 (25.21 /48)

On the road:

JSmoove = 18.8

JJ = 14.8

Al = 14.7

Marvin = 13.6 (19.49/48)

Chillz = 13.0 (21.12 /48)

In Wins:

JSmoove = 23.6

JJ = 22.1

Al = 16.5

Marvin = 17.0 (22.14/48)

Chillz = 15.9 (27.18/48)

In Losses:

JSmoove = 18.2

JJ = 15.0

Al = 15.8

Marvin = 13.7 (19.33/ 48)

Chillz = 13.3 (20.82 /48)

Summary

The summary of all this is that our team losses when JJ has bad games. There's a significant decrease in Smoove and JJ's games in a loss vs. a win. ~6 points. So far, Bibby's efficiency isn't that great. But there again, I would love to see these numbers weighted a little lower for points.

[TheNorthCydeRises]

It's because player efficiency is mainly an offensive stat. It's pretty obvious that this team lives and dies with how JJ is scoring. No other player on the team can really consistently pick up the scoring slack, when he's struggling. That's the difference between our team, and most of the teams who are playoff contenders right now.

[Guest]

Quote:

---

The summary of all this is that our team losses when JJ has bad games. There's a significant decrease in Smoove and JJ's games in a loss vs. a win. ~6 points. So far, Bibby's efficiency isn't that great. But there again, I would love to see these numbers weighted a little lower for points.

They do have stats that have weights on each variable. John Hollinger's PER puts weights on basically the same variables as EFF and Dave Berri's Win Score does the same thing. I prefer Win Score because he regressed each variable over wins to calculate his weights while I believe Hollinger has a more subjective reason behind his weights.

Like all stats, they should be interpretted correctly for them to be useful. Its not like you can take player A and player B on two seperate teams and then look at just their EFF/PER/Win Score to determine the better player. It is more indepth than that. What D is doing is right though, pointing out that in losses Smoove and Joe have a larger differential which more than likely shows that they are better indicators as to how the team will perform (win or lose).

[Diesel]

Quote:

---

he regressed each variable over wins to calculate his weights

Does this mean that a player on a losing team gets a lower score?

What about guys like Al Jefferson?

What about KG last year?

What about Dwayne Wade and Shawn Marion??

[Guest]

No what it means is that he took all the data available to him since I believe 1974 and regressed what the weights for each variable was in relation to wins. Its a statistical method of figuring out basically how many rebounds contribute to a win. So after he has regressed all the variables on wins, he knows the linear weights of what to multiply Rebounds, Field Goals, Assists...and then sum all the variables up. Once you sum up all these variables, it shows you your "win score", meaning how many wins you produced based on his weights on each variable.

So in a nutshell, no guys like Al Jefferson are not penalized for playing on a team that loses. He is, however, somewhat penalized for having crappy players which probably deflate his Win Score by him not accumulating more assists or having to take more shots which lowers his FG%, etc.

If you are interested in this stuff, I suggest reading "Wages of Wins" by Dave Berri. He talks about this statistic more indepth and also talks about other stats in sports. The book also is not very math based, Dave Berri is an economist at some school in Cali so although he uses a lot of econometrics he explains it in real terms instead of heavy math. If you want to see the math, he does have it in the appendix. Big thanks to KB21 for pointing that book out to me, I read it and enjoyed it very much.