Friday, November 24, 2006

A Modest Proposal

There's always some wonk who wants to crunch numbers to find out who's the best player on the court. Today, that happens to be me.

Introducing the Hoosierfunball Player Efficiency Rating (HPER):

Pts+Total Rebs+Off.Rebs.+[2x(Assists+Steals+Blocks)]-[Missed FGs+(2xTurnovers)+(Missed FTs/2)]
all divided by Minutes Per Game=Individual HPER
all divided by Minutes Per Game Possible=Adjusted HPER

Some of this is obvious, but offensive rebounds equal another possession taken away from an opponent, so I want to reward that. Also assists reflect good passing, and passes that lead to free-throws are never recorded, so I double them as well. Steals are also a possession gained (double 'em), and while Blocks certainly are not an indication of gaining a possession, intimidations are rarely counted at the college level. The additional deflections that might show
good defense cannot really be reflected in steals & blocks due to the bad defense that often comes from gambling. For example, at Indiana, Earl Calloway allowed guys like Drew Nietzel to blow by him for lay-ups when he was gambling for the steal. And everyone should remember George Leach's refusal to get defensive rebounds in his pursuit of blocked shots.
Since turnovers are a possession lost, that should penalized above missed shots (double 'em),
while missed free throws only reflect the loss of one point, not a whole possession (halve 'em).

Dividing by minutes played is the standard way to assess the value of the individual productivity (at least, in NBA rating systems). However, one also needs to consider the value to the team, and this is often overlooked in assessing player effectiveness. If someone gets in foul trouble, suspended, injured, technical fouls, all that obviously hurts their ability to produce and help the team.

DJ White: (14ppg/7.5rpg/1apg/1spg/1.8bpg)
{56pts+30total rebs +9off.rebs +[2x(4assists+4steals+7blocks)]
-(18missed fgs+(2x8 turnovers)+(7missed Free throws/2)]}/98minutes played

So, DJ has a .893 Individual HPER (IH).
But here's the important rating for me, Indiana played 4 games, or 160 minutes total that DJ could've played in. While it's never a good idea to play anyone for 40 minutes, it gives a better idea of the per-game productivity and importance of the player to the team:

87.5 divided 160 total possible minutes gives DJ White a .547 Adjusted HPER (AH)

Compare Spencer Tollackson with Courtney Sims, and then Northwestern's Vince Scott, just for fun.
Spencer Tollackson, Minn. (13.3ppg/9rpg/1.7apg/1.7bpg)

101-33=68/74=.919 IH

68/160=.425 AH

Courtney Sims, Mich. (16.6ppg/6.4rpg/1.2bpg)

120/158=.759 IH

120/240=.500 AH


Vince Scott, NU (7.3ppg/2.3rpg/3.8apg/1.5spg/1bpg)

92-36.5=55.5/121=.459 IH
55.5/160=.347 AH

Due to the the difference in competition, I don't think this will be all that meaningful pre-conference, but once we get some conference numbers in play, I'm really looking forward to seeing how this theory shakes out.

Just for fun, here's how Big Ten Opponent's records stack up at this VERY EARLY stage.

OSU Opp.:15-7
UMinn. Opp.: 10-5
Pur. Opp.:12-11
MSU Opp.:16-15
Iowa Opp.:9-8
IU Opp.: 8-8
NU Opp.:9-10
UIUC Opp.:8-13
UMich Opp.:10-18
PSU Opp.:5-12
UW Opp.:3-12


Post a Comment

<< Home