Expected Performance and The Four Factors - Page 2
Old 11-27-2013, 11:18 AM   #21 (permalink)
LOG
is pounding the rock!

Senior Member
 
LOG's Avatar
 
Join Date: Jun 2011
Location: Toronto
Posts: 8,117
Representing:
Default

Someone should send this to MU!
LOG is online now   Boss Key Wife Key Reply With Quote
Old 11-27-2013, 12:54 PM   #22 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,278
Representing:
Default

Quote:
Originally Posted by DanH View Post
Terrence Ross:

Expected Performance:

Value: Team | Opponent
eFG%: 45.2% | 51.8%
ORB%: 28.9% | 24.6%
FT/100: 21.1 | 17.1
TOV%: 12.4% | 14.5%

Actual Performance:

Value: Team | Opponent
eFG%: 47.5% | 45.9%
ORB%: 30.4% | 23.5%
FT/100: 25.5 | 21.1
TOV%: 13.7% | 15.7%

Difference:

Value: Team | Opponent
eFG%: +5.1% | -11.3%
ORB%: +5.1% | -4.6%
FT/100: +20.9% | +23.3%
TOV%: +10.8% | +8.6%

ORTG: 103.5
DRTG: 95.9
Wins: 61

T-Ross the superstar. This shows just how much he's meant to the bench this year - his expected performance is pretty terrible (based on the terrible players he plays with) and so since he actually performs pretty solidly (read: actually pretty awesome eFG% suppression), he just rocks this stat. He gives back a bit with increased fouls, but considering the low eFG% against, high eFG%, and high TOV% for the opposition, this high mark looks well earned.
his +/- are off the charts this season - very impressive, supporting the subjective feel you get when you see him on the court. Some of his numbers maybe skewed because he replaces gay a lot, and gay has been so terrible at times.
moremilk is online now   Boss Key Wife Key Reply With Quote
Old 11-27-2013, 03:54 PM   #23 (permalink)
La Decima!!!!!

Senior Member
 
Join Date: Jan 2012
Location: Toronto, Ontario, Canada
Posts: 7,223
Representing:
Default

COMBINE those stats with data like stop%age = godly stats
LET'S GO RAPTORS!!!!! is offline   Boss Key Wife Key Reply With Quote
Old 11-27-2013, 03:55 PM   #24 (permalink)
La Decima!!!!!

Senior Member
 
Join Date: Jan 2012
Location: Toronto, Ontario, Canada
Posts: 7,223
Representing:
Default

Quote:
Originally Posted by LOG View Post
Someone should send this to MU!
I imagine MU has a stats and logistics division - or at least one guy
LET'S GO RAPTORS!!!!! is offline   Boss Key Wife Key Reply With Quote
Old 11-28-2013, 11:01 AM   #25 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,125
Representing:
Default

Quick updated win estimates:

I decided to try to reduce impacts by half - originally applied 40-25-20-15 weight to offensive impact and 40-25-20-15 weight to defensive impact. Strikes me now that I should halve that - apply 40-25-20-15 to entire thing, ie 20-12.5-10-7.5 to offense and to defense. This yields the following results:

Player | ORTG | DRTG | Wins
DD | 101.4 | 104.3 | 33
Gay | 96.9 | 101.5 | 28
Lowry | 100.4 | 99.2 | 44
Amir | 101.4 | 96.3 | 55
Jonas | 94.3 | 98.6 | 29
Ross | 101.7 | 97.9 | 52
TH | 99.6 | 99.3 | 42
Fields | 101.2 | 103.1 | 36

Predicted win %: .476 | 6.2 - 6.8 after 13 games | 39 Wins Predicted

Still holds up, gets rid of the extreme results. Each player's value sits a little closer to what you'd expect.
DanH is offline   Boss Key Wife Key Reply With Quote
Old 11-28-2013, 11:06 AM   #26 (permalink)
La Decima!!!!!

Senior Member
 
Join Date: Jan 2012
Location: Toronto, Ontario, Canada
Posts: 7,223
Representing:
Default

Quote:
Originally Posted by DanH View Post
Quick updated win estimates:

I decided to try to reduce impacts by half - originally applied 40-25-20-15 weight to offensive impact and 40-25-20-15 weight to defensive impact. Strikes me now that I should halve that - apply 40-25-20-15 to entire thing, ie 20-12.5-10-7.5 to offense and to defense. This yields the following results:

Player | ORTG | DRTG | Wins
DD | 101.4 | 104.3 | 33
Gay | 96.9 | 101.5 | 28
Lowry | 100.4 | 99.2 | 44
Amir | 101.4 | 96.3 | 55
Jonas | 94.3 | 98.6 | 29
Ross | 101.7 | 97.9 | 52
TH | 99.6 | 99.3 | 42
Fields | 101.2 | 103.1 | 36

Predicted win %: .476 | 6.2 - 6.8 after 13 games | 39 Wins Predicted

Still holds up, gets rid of the extreme results. Each player's value sits a little closer to what you'd expect.
Once again - just furthers my position to KL and Amir's importance to the team
LET'S GO RAPTORS!!!!! is offline   Boss Key Wife Key Reply With Quote
Old 11-28-2013, 01:15 PM   #27 (permalink)
whatever

Senior Member
 
Join Date: Dec 2007
Location: toronto
Posts: 9,278
Representing:
Default

Quote:
Originally Posted by DanH View Post
Quick updated win estimates:

I decided to try to reduce impacts by half - originally applied 40-25-20-15 weight to offensive impact and 40-25-20-15 weight to defensive impact. Strikes me now that I should halve that - apply 40-25-20-15 to entire thing, ie 20-12.5-10-7.5 to offense and to defense. This yields the following results:

Player | ORTG | DRTG | Wins
DD | 101.4 | 104.3 | 33
Gay | 96.9 | 101.5 | 28
Lowry | 100.4 | 99.2 | 44
Amir | 101.4 | 96.3 | 55
Jonas | 94.3 | 98.6 | 29
Ross | 101.7 | 97.9 | 52
TH | 99.6 | 99.3 | 42
Fields | 101.2 | 103.1 | 36

Predicted win %: .476 | 6.2 - 6.8 after 13 games | 39 Wins Predicted

Still holds up, gets rid of the extreme results. Each player's value sits a little closer to what you'd expect.
couple of ideas that may help, one is to get rid of FT% because it's already captured (indirectly) in the eFG. Either that, or vastly reduce the percentage for FT.

second, use the numbers from the first calculation and do a second round where you normalize the rating as if that player would play an equal number of minutes with any other teammate. In other words, say demar has a defensive rating of 104. If Lowry plays 20 minutes with demar and gay plays 30 minutes, demar's poor defense impacts lowry's numbers a lot less than gay's. So you can do some kind of normalization. I think this number would be more accurate in capturing one's impact while minimizing the impact of teammates, no?
moremilk is online now   Boss Key Wife Key Reply With Quote
Old 11-28-2013, 02:38 PM   #28 (permalink)
sticking to his guns... is stuck to his guns...

Senior Member

 
DanH's Avatar
 
Join Date: Jul 2010
Location: Montreal, QC
Posts: 4,125
Representing:
Default

Quote:
Originally Posted by moremilk View Post
couple of ideas that may help, one is to get rid of FT% because it's already captured (indirectly) in the eFG. Either that, or vastly reduce the percentage for FT.

second, use the numbers from the first calculation and do a second round where you normalize the rating as if that player would play an equal number of minutes with any other teammate. In other words, say demar has a defensive rating of 104. If Lowry plays 20 minutes with demar and gay plays 30 minutes, demar's poor defense impacts lowry's numbers a lot less than gay's. So you can do some kind of normalization. I think this number would be more accurate in capturing one's impact while minimizing the impact of teammates, no?
eFG% does not include FT%. TS% does. The 4 factors regression was done with these four factors - I'm not sure how changing to TS% and eliminating FT would impact the correlation with wins.

Not sure what you mean with the second request. You are suggesting I should create a second set of expectations, based on players' relative performances rather than their raw ones? Then calculate a new relative performance based on those new expectations. Could you not then make the argument that you should do that again? And again? And again into forever?

Hmm. I'll look into it and see what the impact is. In spite of my best efforts, you've intrigued me. Not sure how I'll do it though.
DanH is offline   Boss Key Wife Key Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


All times are GMT -4. The time now is 09:37 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright RaptorsForum.com 2005-2011

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24