Formula for value as a playmaker with some results

Moderator: Doctor MJ

Lou Fan
Pro Prospect
Posts: 790
And1: 711
Joined: Jul 21, 2017
     

Formula for value as a playmaker with some results 

Post#1 » by Lou Fan » Thu Sep 7, 2017 4:41 am

(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) * (((ast / tov) * 0.8) + a * 0.2) * ((mpg + 72.0) / 120.0))
a= assists per 100
usg= usage rate
ast= assist %
tov= turnover %
mpg= minutes per game
I know this formula looks insane but everything has a purpose.
(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) This part determines how much you have the ball aka your opportunity to make an assist people who have less opportunity but same assists get rewarded.
((ast / tov) * 0.8) This part rewards people who have good assist to turnover ratios pretty straightforward.
+ a * 0.2) This rewards volume
((mpg + 72.0) / 120.0)) This rewards people who do it over more minutes giving credit here from trex I borrowed this from his scoring rating.
Some Results:
Stockton 89-90:1.2
Westbrook 16-17: .71
Paul 07-08: 1.2
Bird 86-87: .67
Nash 04-05: .86
Kidd 01-02: .87
C-Webb 02-03: .51
Kobe 07-08: .50
Magic: 1.02
Though I'm pretty happy with these obviously it needs some work. Any ideas or advice would be much appreciated. Especially advise concerning how to make big-man passers numbers better. Thanks!
Edit: Maybe it's only good for comparing across positions, but I'd rather it be good for all positions.
smartyz456 wrote:Duncan would be a better defending jahlil okafor in todays nba
euroleague
General Manager
Posts: 8,448
And1: 1,871
Joined: Mar 26, 2014
 

Re: Formula for value as a playmaker with some results 

Post#2 » by euroleague » Fri Sep 8, 2017 7:39 am

twolves97 wrote:(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) * (((ast / tov) * 0.8) + a * 0.2) * ((mpg + 72.0) / 120.0))
a= assists per 100
usg= usage rate
ast= assist %
tov= turnover %
mpg= minutes per game
I know this formula looks insane but everything has a purpose.
(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) This part determines how much you have the ball aka your opportunity to make an assist people who have less opportunity but same assists get rewarded.
((ast / tov) * 0.8) This part rewards people who have good assist to turnover ratios pretty straightforward.
+ a * 0.2) This rewards volume
((mpg + 72.0) / 120.0)) This rewards people who do it over more minutes giving credit here from trex I borrowed this from his scoring rating.
Some Results:
Stockton 89-90:1.2
Westbrook 16-17: .71
Paul 07-08: 1.2
Bird 86-87: .67
Nash 04-05: .86
Kidd 01-02: .87
C-Webb 02-03: .51
Kobe 07-08: .50
Magic: 1.02
Though I'm pretty happy with these obviously it needs some work. Any ideas or advice would be much appreciated. Especially advise concerning how to make big-man passers numbers better. Thanks!
Edit: Maybe it's only good for comparing across positions, but I'd rather it be good for all positions.


It rewards ball-hogs. Especailly ball hogs who can't shoot. Having a high assist% and not shooting the ball yourself will (based off assists per 100 and assist % being high) still benefit you based off of your USG corrector.

Try Rondo in this formula from 2012.

Also should correct for league averages, as they change across eras.
Lou Fan
Pro Prospect
Posts: 790
And1: 711
Joined: Jul 21, 2017
     

Re: Formula for value as a playmaker with some results 

Post#3 » by Lou Fan » Sat Sep 9, 2017 3:03 am

euroleague wrote:
twolves97 wrote:(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) * (((ast / tov) * 0.8) + a * 0.2) * ((mpg + 72.0) / 120.0))
a= assists per 100
usg= usage rate
ast= assist %
tov= turnover %
mpg= minutes per game
I know this formula looks insane but everything has a purpose.
(((((a * 1.98 / 100.0) + usg) ** 0.5) ** -1.0) This part determines how much you have the ball aka your opportunity to make an assist people who have less opportunity but same assists get rewarded.
((ast / tov) * 0.8) This part rewards people who have good assist to turnover ratios pretty straightforward.
+ a * 0.2) This rewards volume
((mpg + 72.0) / 120.0)) This rewards people who do it over more minutes giving credit here from trex I borrowed this from his scoring rating.
Some Results:
Stockton 89-90:1.2
Westbrook 16-17: .71
Paul 07-08: 1.2
Bird 86-87: .67
Nash 04-05: .86
Kidd 01-02: .87
C-Webb 02-03: .51
Kobe 07-08: .50
Magic: 1.02
Though I'm pretty happy with these obviously it needs some work. Any ideas or advice would be much appreciated. Especially advise concerning how to make big-man passers numbers better. Thanks!
Edit: Maybe it's only good for comparing across positions, but I'd rather it be good for all positions.


It rewards ball-hogs. Especailly ball hogs who can't shoot. Having a high assist% and not shooting the ball yourself will (based off assists per 100 and assist % being high) still benefit you based off of your USG corrector.

Try Rondo in this formula from 2012.

Also should correct for league averages, as they change across eras.

You're totally right I have since adjusted the formula to make the usage control way more but if you have any suggestions (literally at all) I'd love to hear them.
Rondo is 11-12 .64
CP3 07-08 is .81
Magic 85-86 is .63
So yeah still clearly needs some work and I have been wracking my brain on how to fix it and frankly I'm stumped. I was thinking of adding an ORTG-scoring factor but that wouldn't work as it would make Shaq look good in my metric because of his gravity. :banghead: :banghead: So please guys if you have any ideas chime in.
smartyz456 wrote:Duncan would be a better defending jahlil okafor in todays nba

Return to Statistical Analysis