When my colleague John Gasaway made his pick for national Player of the Year last week, one thing was conspicuous by its absence in his column--overall player ratings. For those of us more familiar with statistical analysis in the NBA, where player rating systems often dominate the conversation, this seems a little strange.
In the NCAA, most analytical work by Gasaway, Ken Pomeroy and others has focused at the team level and individual player skills. There is much to be said for this line of thinking. Until we create the elusive (and probably mythical) "Holy Grail" system, player ratings will suffer from the difficulty of contextualizing how players fit together as well as the issue of accounting for plays which take place outside the scope of the box score. Still, overall player ratings can help us understand how to weight players' advantages in various categories. Does DeJuan Blair's superior offensive rebounding make up for Blake Griffin's larger role in the Oklahoma offense? That's difficult to judge subjectively.
There's nothing inherently stopping NBA value metrics from being ported to an NCAA system. DraftExpress.com calculates PERs for college players, for example. Of course, six of the top 10 players in their rankings come from non-BCS conferences, which illustrates the danger in using these numbers without any adjustment for strength of schedule.
Here, I think the Pelton NCAA-to-NBA Translations, which I introduced on this site last June and used to evaluate top returning prospects in College Basketball Prospectus 2008-2009, offer a unique perspective. My system uses the transition of various players from the college game to the pros to empirically weight strength of schedule. While the differing adjustments applied to each statistical skill (two-point percentage, offensive rebound percentage, etc.) is a factor, in some sense the translations can be considered a schedule-adjusted measure of player performance using my WARP rating system.
So, using data provided by Pomeroy, I crunched the numbers for this year's group of players through the conference tournaments. At the top, the result was hardly surprising. Blair came out as the top player, backing up Gasaway's reading of the numbers. What was unexpected was the magnitude of that margin:
Player Win% ORtg DRtg
Blair .721 110.7 103.4
Griffin .597 106.5 103.6
Griffin's translated .597 winning percentage (estimating that a rookie Griffin plus four average teammates would win nearly 60 percent of their games) is impressive. It's slightly better than what Michael Beasley achieved last year (.594), though behind Kevin Love's .645 translated winning percentage. As Gasaway said, many years Griffin would have been the clear choice for Player of the Year. Alas, at the same time Griffin was stealing headlines, Blair was putting together a season that was historic in its efficiency. The translations point to the importance of his turnover-free style; Blair came out slightly ahead on offense even before his off-the-charts offensive rebounding was added to the mix.
That debate considered, here are a handful of other observations from the translations.
The Imaginary All-American Team
G Ty Lawson, North Carolina (.541 Win%)
G James Harden, Arizona State (.504)
F Trevor Booker, Clemson (.552)
Booker is the only real surprise of the group, earning a spot on the strength of leading the ACC in both field-goal percentage and rebounding. He's also a contributor in both steals (1.5 per game) and blocks (2.0 per game), something my system particularly tends to like.
- Stephen Curry wasn't particularly close. The Davidson sensation comes out with the highest translated usage rate in the country, having been a one-man show all season long. His passing also improved as he made the transition to the point. Overall, however, my numbers show Curry to have been more effective playing off the ball as a sophomore. His translated True Shooting Percentage shrunk from 51.4 percent to 47.7 percent. The All-America honors Curry will likely get are something of a stretch.
- John Bryant was the nation's best mid-major player. You may know Bryant as the guy amongst the nation's rebounding leaders who is neither Blair nor Griffin. Bryant led the country in defensive rebound percentage (36.3 percent) and was second to Blair on the offensive glass (18.6 percent). Thanks to that and 59.6 percent shooting from the field, the 300-pounder actually comes out as the NCAA's third-best performer after Blair and Griffin. Adjusting for schedule strength knocks all but two mid-major players out of the top 20 in the rankings, but Bryant still comes out looking like an elite player.
- Greg Monroe was the best freshman in the country. No surprise there, as the highly-touted Monroe mostly lived up to the hype on a Georgetown team that failed to give him enough support. Here, the numbers confirm the notion that it was a down year for diaper dandies. Monroe (.514 translated winning percentage) was tops amongst freshmen by a mile, a dramatic change from last year, when two of the top three performers were one-and-done standouts Beasley and Love.
Kevin Pelton is an author of Basketball Prospectus.
You can contact Kevin by clicking here or click here to see Kevin's other articles.