It explains about 60-70% of the variation depending on what data sources you use. However, you have to assume that recruit rankings and on the field success are independent variables. They aren’t, which is why Harrison Bailey isn’t a five star, but if Georgia offers a kid with a fast forty time he becomes an automatic four star.
After spending several years running numbers on this stuff, I think recruit rankings are f@&$ed. Who is evaluating the evaluators? Who has more skin in the game? Barton Simmons who needs clicks for revenue? Or Jeremy Pruitt who needs to win ball games?
I trust the dude who’s put dudes into the NFL, versus the ranking services who were exposed when they ranked a player that didn’t even exist, and are clearly on the take when going to one of their camps gets you a rankings boost.
Let me start by saying I agree because this might sound like a bit of a ramble.
The recruiting services evaluate high school talent to get subscriptions to their services. They know the most likely subscribers are those from "big" schools. They skew their numbers that way. That's not to say they are automatically wrong, but their driving force is making money.
In fact, they are often right (as the numbers kind of show), but I suspect they are correct more so because talented kids are going to do better where they are surrounded with other talented kids and immersed in a culture that drives them toward success. In other words, they are going to succeed more often at "big" schools that emphasize football. Therefore, the players that commit to those schools will get higher ratings because they are more likely to succeed. This is important because recruiting services want to be right so they get more subscriptions.
Now, that doesn't mean it's automatically skewed. Big, strong, fast, kids are easy to spot. The tricky part is determining which ones will be the NFL-level players. Hits and misses are all over the place, but the biggest, fastest, and strongest are more likely to hit than the other ones. If it was that easy, they'd all agree and always be right.
That's why I think recruiting rankings are more reactive than they are predictive. They aren't unimportant, but that aren't all-important either. What you really want is to be a school that drives services to bump your commits because you're more likely to give them success than most other schools.
