Posted: Tue May 06, 2008 12:47 pm Post subject: Pace Factor, Expected Wins, and the PER
Hello Everyone,
I really enjoy this board and its pursuit of statistical diversion from the basketball norms. I have three questions regarding the above topics.
Regarding the Pace factor and expected wins, I know the formulas, but I'm not sure of the results. Here are a couple of examples using data from a couple of leagues that I'm in.
1) Detroit Pistons (71-72) 27-47
Pace Factor = (Team Poss + Opp Poss X 48) / Team Minutes X 2
Pace Factor = ((8552+8537) X 48) /17794 X 2
Pace Factor = 820272 / 35588
Pace Factor = 23.4
**23.4 points IS NOT my team's pace for that particular season. What am I doing wrong?
Altoona = 82 X (10228^16.5) / (10228^16.5 + 7858^16.5)
Altoona = 82 X (1.45^16.5) / (1.45 X 10^66 + 1.87 X 10^64)
Altoona = 82 X (145 / 145 + 18.7)
Altoona = 82 X (145/163.7)
Altoona = 82 X .885
Altoona = 72.57
Altoona = 73 Wins
Please let me know if this is correct, or, where did I go wrong?
3) PER-- Does anyone have an entry formula for Hollinger's PER in Excel format? I thought that I'd ask before I struggle through the task of constructing one.
Thanks,
Miguel _________________ LOOK OUT BELOW!!! WITH NO REGARD FOR HUMAN LIFE!!
Joined: 13 Oct 2005 Posts: 530 Location: Atlanta, GA
Posted: Tue May 06, 2008 6:17 pm Post subject:
1. "Team Minutes" needs to be divided by 5 to account for the 5 players on the court at any given time.
2. Expected winning % is just found by: (points scored^16.5)/((points scored^16.5)+(points allowed^16.5)). So if I'm calculating it for the 2008 Celtics, who scored 8245 pts and allowed 7404, I would plug in: (8245^16.5)/((8245^16.5)+(7404^16.5)) = .855. Multiply .855 by 82, and you get 70 expected wins. 14 is a better exponent than 16.5, though.
3. I just posted a PER excel worksheet to the APBRmetrics files Yahoo group. It's from 2006 (the last year I actually bothered to calculate PER), but you can apply it to any season you want.
1. "Team Minutes" needs to be divided by 5 to account for the 5 players on the court at any given time.
2. Expected winning % is just found by: (points scored^16.5)/((points scored^16.5)+(points allowed^16.5)). So if I'm calculating it for the 2008 Celtics, who scored 8245 pts and allowed 7404, I would plug in: (8245^16.5)/((8245^16.5)+(7404^16.5)) = .855. Multiply .855 by 82, and you get 70 expected wins. 14 is a better exponent than 16.5, though.
3. I just posted a PER excel worksheet to the APBRmetrics files Yahoo group. It's from 2006 (the last year I actually bothered to calculate PER), but you can apply it to any season you want.
In following with John Hollinger's most recent book, "Pro Basketball Forecast, 2005-06," he adjusted his Expected Wins formula by raising the power from 14 to 16.5.
I was simply doing it from that perspective. It's become painfully obvious, though, that 16.5 is simply too big a number to work with. Too much reduction involved. Fourteen is basically a more even value that returns closer to round results. It also seems to be a more universally accepted standard.
Another proof of this is when you put the formulas in Excel.
BTW, thanks for the explanation wylie of the division by 5 in Pace. Now that I've conquered that hurdle, now it's on to League VOP.
Thanks guys, for the input. feel free to share more.
Miguel _________________ LOOK OUT BELOW!!! WITH NO REGARD FOR HUMAN LIFE!!
Joined: 13 Oct 2005 Posts: 530 Location: Atlanta, GA
Posted: Thu May 08, 2008 2:01 pm Post subject:
dudeimstoked wrote:
davis21wylie2121 wrote:
1. "Team Minutes" needs to be divided by 5 to account for the 5 players on the court at any given time.
2. Expected winning % is just found by: (points scored^16.5)/((points scored^16.5)+(points allowed^16.5)). So if I'm calculating it for the 2008 Celtics, who scored 8245 pts and allowed 7404, I would plug in: (8245^16.5)/((8245^16.5)+(7404^16.5)) = .855. Multiply .855 by 82, and you get 70 expected wins. 14 is a better exponent than 16.5, though.
3. I just posted a PER excel worksheet to the APBRmetrics files Yahoo group. It's from 2006 (the last year I actually bothered to calculate PER), but you can apply it to any season you want.
Any reason you use 16.5 as the exponent rather than 14?
When I calculate pyth, I never use anything except 14. For the purposes of his example, though, I kept it at 16.5, since he was already using it and I didn't want to add to the confusion. But for future reference, always use 14.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum