urtube suggested this vidya to me, small creator, pretty well put together. A decent bit of the info he goes over more or less confirms conventional wisdom around recruiting.
He broke down the data in a few different ways. He used pro bowl selections, the nfl draft, draft position, and national championships. The vidya is a bit more focused on the predictive power of recruiting rankings re: the NFL, but there is some usefulness for college also imo.
Some quick highlights and then some sperging further down:
1) at the all-pro selection level, recruiting rankings aren’t very useful at all,
2) 58% of 5 star recruits were drafted, 20% of 4 stars were drafted. A 2 or 3 star ranking was more or less indicative of an FBS level player but not generally expected to go to the NFL.
3) all 23 national champions in the recruiting era had at least one top 10 class on their roster.
4) QB, OL, MLB are the hardest positions to predict. DL, DE/edge, RB, WR are the easiest to predict.
5) CB is harder to predict but he hypothesizes that it’s because the best HS kids aren’t put at corner, they’re put at LB or safety where they can be around the ball more and then converted later.
Overall I think a lot of his analysis can be applied to college and as I said at the beginning of the post, it more or less backs up a lot of the conventional wisdom around recruiting, especially the importance of defensive line talent in college football. He talked about offensive line talent being harder to evaluate, said it was more important to get kids to camp and evaluate in person. That was interesting also, that goes a bit more against my personal assumptions but that’s fine. I guess the idea is that it’s a more cerebral position than I would’ve assumed.
To that point: he touched on this and I more or less agree - the more tangible the required skills are for a position, the easier it is to evaluate and predict. Edge rushers need to be big and fast so a freak athlete is easy to evaluate. Versus a QB or MLB where a lot more of their performance is mental and that’s just difficult to predict years down the line.
One of his main conclusions was that the further you get from high school, the less predictive recruiting rankings are, which makes sense. 58% of 5 stars getting drafted is high - and in college, if you have 5 star kids on your roster that kid is going to be a big contributor for you most of the time. But by the time you get to the pro bowl selections, 5 stars make up only 13% of selections. The data on pro bowlers isn’t a huge surprise imo but still interesting - it’s actually kind of wild that 60% of pro bowlers were 3 star or less. And again, a lot of these are in the positions that are harder to predict.
FUN