DSMok1
Member
- Joined
- Nov 10, 2008
- Messages
- 340
- Reaction score
- 0
For all you stat geeks, I have completed the application of the statistical plus/minus formula to the NCAA stats from last year. What does this do? It provides perhaps the best look at who the best players actually were from last year. Or more, what players actually helped the most.
"Statistical Plus/Minus" attempts to derive what a player's adjusted plus/minus was from the box score stats. It is less noisy than the APM itself, but does miss somewhat on some player's defensive contributions. In the NCAA, stellar players from secondary conferences can be a little overrated (though this actually is measuring their production--they produced a lot against their weak foes).
This thread discusses this work over at the APBRmetrics forum: http://sonicscentral.com/apbrmetrics/viewtopic.php?t=2368
Basically, I took the NBA formula and used it for college, and then applied the appropriate adjustment factor to account for strength of schedule.
The top players from last year by 95% confidence level (the level above which we can be 95% confident they were--this combines actual production with minutes played.):
And for anyone interested, here are the top returning players in the Big XII, by 95% confidence level again:
Oddly enough, Willie isn't showing up very well... according to these stats, he was 4th on the team last year.
He was a freshman, though, and the freshman-sophomore transition should bump his results quite a bit.
Look at the full Google Spreadsheet if you'd like. There are tons of things you can do with it.
Likely breakout players:
BJ Holmes
Jamar Samuels
Dexter Pittman
Jamie Vanderbeken
Keith Ramsey
Laurence Bowers
"Statistical Plus/Minus" attempts to derive what a player's adjusted plus/minus was from the box score stats. It is less noisy than the APM itself, but does miss somewhat on some player's defensive contributions. In the NCAA, stellar players from secondary conferences can be a little overrated (though this actually is measuring their production--they produced a lot against their weak foes).
This thread discusses this work over at the APBRmetrics forum: http://sonicscentral.com/apbrmetrics/viewtopic.php?t=2368
Basically, I took the NBA formula and used it for college, and then applied the appropriate adjustment factor to account for strength of schedule.
The top players from last year by 95% confidence level (the level above which we can be 95% confident they were--this combines actual production with minutes played.):
Code:
1 Stephen Curry
2 DeJuan Blair
3 Ty Lawson
4 Blake Griffin
5 Lester Hudson
6 Terrence Williams
7 James Harden
8 Toney Douglas
9 Eric Maynor
10 Danny Green
11 Tony Gaffney
12 Talor Battle
13 Kenneth Faried
14 Nick Calathes
15 Gordon Hayward
16 Jermaine Taylor
17 Tywain McKee
18 Kevin Palmer
19 Manny Harris
20 Jared Quayle
21 Taj Gibson
22 Evan Turner
23 Kyle Singler
24 Robbie Hummel
25 Marqus Blakely
And for anyone interested, here are the top returning players in the Big XII, by 95% confidence level again:
Code:
1 Cole Aldrich Kansas
2 Cory Higgins Colorado
3 Damion James Texas
4 James Anderson Oklahoma St.
5 B.J. Holmes Texas A&M
6 J.T. Tiller Missouri
7 Sherron Collins Kansas
8 Ryan Anderson Nebraska
9 Craig Brackins Iowa St.
10 Jamar Samuels Kansas St.
11 Brady Morningst Kansas
12 LaceDarius Dunn Baylor
13 Zaire Taylor Missouri
14 Willie Warren Oklahoma
15 Tweety Carter Baylor
16 Jacob Pullen Kansas St.
17 John Roberson Texas Tech
18 Obi Muonelo Oklahoma St.
19 Justin Mason Texas
20 Dexter Pittman Texas
21 Ron Anderson Kansas St.
22 Bryan Davis Texas A&M
23 Donald Sloan Texas A&M
24 Marcus Morris Kansas
25 Mike Singletary Texas Tech
Oddly enough, Willie isn't showing up very well... according to these stats, he was 4th on the team last year.
He was a freshman, though, and the freshman-sophomore transition should bump his results quite a bit.
Look at the full Google Spreadsheet if you'd like. There are tons of things you can do with it.
Likely breakout players:
BJ Holmes
Jamar Samuels
Dexter Pittman
Jamie Vanderbeken
Keith Ramsey
Laurence Bowers