WARP....just for fun

[RaceToTheTop 23,603]

I mentioned this in another thread, buut at one point I had worked on creating a "WARP" (wins above replacement player for those not into baseball analytics) for college basketball players.  I'm in the process of rebuilding it and running it for IU (and other players).  

Here is the jist of it:  WARP is meant to reflect a player's value in relation to a replacement level NCAA Division 1 player.  I define that player to be an average player playing at a low division 1 school;  a team consisting of such players, playing against average division 1 opponents, I give estimate values of 57 points scored per game (40% shooting on 60 shots, 60% free throw shooting on 15 free throws per game, 20 turnovers per games, assists on half of all scored goals, 5 steals, 1.5 blocks per game as a team).  As part of the WARP calculation, I refigure the team's stats based, replacing a replacement player with the player I am calculating value for.  I then figure a new scoring average based on what that player adds.  Unfortunately defense is hard to figure, so the best I can do is use the player's steal and blocked shot stats to increase or decrease their value;  this number is added or subtracted to their offensive value since the number I am creating is meant to be an overall value.

I've ran the current calculations for IU players....note that the following numbers do not factor in strength of schedule.  Additionally, the numbers posted are the per/40 numbers;  i.e., the addition or subtraction of value is based on the player playing a full 40 minutes in a game.  Later I will run numbers on WARP, which does adjust for SOS, pace of play, and factors in how much they actually player per game.  The numbers listed below are:

base points (57;  see above) + scoring addition/loss over replacement (based on field goal shooting/rate of shooting/free throw shooting/rate of shooting) + other factors (rebounding/assists/steals/blocks) = new team point value with player

Trayce Jackson Davis:  57 + 8.7 scoring addition + 5.7 other addition = 71.4

Thompson:  57 + 3.6 + 7.1 = 67.7

Durham:  57 + 6.0 + 0.5 = 63.5

Green:  57 + 4.5 + 1.3 = 62.8

Phinissee 57 + 2.2 + 3.0 = 62.2

Brunk  57 + 2.4 + 2.8 = 62.2

Smith 57 + 4.2 + 0.9 = 62.1

Davis 57 + 1.8 + 1.5 = 60.3

Hunter 57 + 2.1 - 0.8 = 58.3

Franklin 57 + 0.7 +0.6 =58.3

I was a little surprised to see Smith that low....and I know it doesn't factor in defense well.  But examining closer, he's a guy who is basically an in the perimeter guy shooting 48%, doesn't shoot free throws well and turns it over a lot.

 

[RaceToTheTop 23,603]

For comparison, here are the ratings for Maryland and Purdue (again, not adjusted for PACE or SOS).

Maryland:

J.Smith  57 + 8.8 + 6.1 = 71.9

Cowan 57 + 8.6 + 3.8 = 69.4

Lindo 57 + 3.8 + 3.2 = 64

Morsell 57 + 3.4 + 2.3 = 62.7

Scott 57 + 2.2 + 2.6 = 61.8

Wiggins 57 + 2.6 + 2.0 = 61.5

Ayala 57 + 1.7 + 0.7 = 59.4

S.Mith 57 - 3.2 + 0.2 = 54.0

PURDUE
Williams 57 + 5.1 + 6.8 = 69.4

Haarms 57 + 5.6 + 4.1 = 66.7

Boudreaux 57 + 3.6 + 3.5 = 64.1

Proctor 57 + 3.4 + 1.2 = 61.6

Hunter 57 + 2.7 + 1.3 = 60.9

Stefanovic 57 + 3.4 +0.5 =60.9

Eastern 57 - 0.6 + 4.9 = 60.4

Wheeler 57 - 1.5 + 3.7 = 59.2

Thompson 57 + 1.1 -1.9 = 56.2

 

B