Tuesday, July 10, 2007

Who Knows What They Are Talking About?

I stumbled upon this link- and found it sort of interesting. It concerns evaluating one of the bigger recruiting websites: Rivals.

The link below is basically the Rivals Top 50 from 2002- the five year class that came out in this current draft- and how their projections faired. Interesting that their top 2 players were total studs in college. Of the entire 50, my off the cuff count is 20 or so were good or great players- another six to eight probably would have contributed at a place like Tulane- and a dozen were hurt or had off the field issues. Particularly if you discount the criminals and dopes- Rivals isn’t really projecting who would be likely to go to, say, Baton Rouge, to commit gun crimes- at least half were good players. Another quarter arguably belonged in I-A- but maybe not a BCS League.

This is pretty powerful evidence that- at least concerning the top players in the country- Rivals is a solid evaluator of potential. I know I didn't expect- after taking out non-football factors- they'd have a strike rate arguably in excess of 60%. While the author clearly has an axe to grind versus the Seminoles, I can agree with his conjecture that more of these guys than not were above average college starters.

NFL teams simply don’t get close to that sort of success ratio in the first 50 guys taken in the draft. It isn't a perfect comparison- but 2/3rds or so of the first 50 NFL picks don't go on to be above average starters (see Patrick Ramsey!)- but Rivals has that hit rate with their first 50 selections- at least in 2002 anyway. At the very least, the list simply isn't guys on a list because LSU is interested in them and not Tulane.

Personally, would guess that Rivals has enough candlepower to at least evaluate the top players. Past that... I dunno; it isn’t definitive- one year is not a sufficient sample size- just interesting.