Q&A: Star rankings vs collegiate success

Time for some viewer mail:

Q: Just wondering if you honestly believe that a high star rating equals out to a succesfull[sic] recruit? If you did any studying at all instead of being just a plain idiot you would realize that players that are 2 or 3 stars typically have more success at the college and NFL level than those "precious" 5 stars that UGAG loads up on every year.

- Anonymous

Yes, Mr. Anonymous. Yes I do believe a high star rating is significant in predicting future success.

Yours is the characteristic song of a Gailey-Johnson slurper in the wild. Though you fail to cite any hard data, I'm aware of the primitive argument you're trying to make, and you do the same thing as the rest: ignore the base number of 4 and 5 star rated players. We've argued against rah-rah jocksniffers like you countless times since before Johnson, and we are frankly tired of doing so, as those of your ilk appear refractory to the accumulation of knowledge. It just so happens however that Popular Yahoo Sports blogger Dr. Saturday is in the midst of a week-long series on this very topic, and he has done all the necessary research for us. It is among the most groundbreaking yet on the subject.

His first opus, on the correlation of star-rankings with future All-American recognition, is most revealing. Here is what Dr. Saturday had to say to those such as yourself:

If you didn't know any better, you might be convinced all those recruiting stars everyone gets worked up about every winter didn't correspond to future success at all (...) Fortunately, because we've been bestowed by the American education system with the magic of basic arithmetic, we do know better. (...) I would hope that two and three-star players could acquit themselves well enough to produce a large number of big names, since they account for more than 85 percent of signees nationally.

As it turns out, going by Rivals.com data, over 10,000 players were awarded either 2 or 3 star rankings in the last 5 recruiting years, while a mere 171 were awarded 5 star ratings and 1,805 were awarded 4 star ratings. Referencing this information against the last 5 All-America teams named by the American Football Coaches' Association, the Associated Press, the Football Writers of America, the Sporting News, and the Walter Camp Foundation, Dr. Saturday calculated the odds for being named to an All-American team based on Rivals star ranking:

5-star: 1 in 13
4-star: 1 in 53
3-star: 1 in 172
2-star: 1 in 744

He concludes(emphasis in bold is ours):

... if you consider the initial [star rating] as a kind of investment – a projection of the how likely a player is of becoming an elite contributor compared to rest of the field – well, you'd put your money with the "experts" over the chances of finding the proverbial diamond in the rough every time... (...) Of course, a large number of players in that sample size haven't finished their careers, but you can divide up the numbers over any time period you'd like – one year, five years, 10 years: The ratio always looks identical on a per-capita basis, and it is not a crapshoot. Four and five-star players are roughly seven times as likely as two and three-star players to land on an All-America team, and the numbers in the NFL Draft tend to be even even more lopsided toward the hyped recruits. All the more reason to want as many of them as you can get your hands on.

To take it even further, long-time Bbuzzoff poster Tech Head went as far as to derive the ratio of All-ACC team selections(1st, 2nd, and Honorable Mention) in 2010 as it relates to the total of ACC players recruited per star category(via Rivals.com) from 2007-2010, excluding kickers. He found that 27.7% of the total number of 5 star players, 9.5% of the 4 stars, 4.2% of the 3 stars, and 2.7% of the 2 stars made the 2010 All-ACC team in some capacity. According to his statistics, this is the 6th straight year such an alignment can be found, which we can only assume to be the entire period over which he has tracked the data.

Moving back to Dr. Saturday though-- he continues in a second article to illustrate the connection between consistently high recruiting rankings with multiple top 10 finishes in the polls, which helps weed out fluke occurrences such as Kansas being in the final top 10 in 2007. In a third article, he even groups teams by recruiting star caliber and compares their records head-to-head against one another. In both instances, the recruiting rankings, while most certainly not fail-safe, were a significant statistical predictor of team success.

Are there exceptions and do the recruiting services get it wrong some of the time? Yes. But I'm sure even The Arrogant One himself has "misevaluated" a player or two in his day. As Dr. Saturday says, I'd much rather have a team loaded with 4 and 5 star players than one loaded with 2 and 3 star players, regardless of how many "geniuses" I have on the coaching staff. On top of that, it's not like all the mid-major fare Coach Johnson has been settling for were his top choices either-- he was going for the 4 and 5 star "UGAG" guys just like everyone else, but he whiffed on them. There is truly no sugar-coating the fact that much of the 2011 class is second-tier even in Johnson's eyes. Of course that should be immediately apparent to anyone who sees offer lists replete with Sun Belt and FCS schools, but the Kool-Aid drinker is hardly one to be rational.

As for UGA, do I need point out to you the fact that their "overrated" 4/5 star recruits have won 9 out of the last 10 head to head against all the 2/3 star "under the radar" guys at Georgia Tech? And 16 of the last 20? How's that for statistical correlation?

But please continue on in your ignorance Mr. Anonymous. By all means do a somersault and wave your pom-poms in support of Coach Johnson's incompetence. We will continue to expose him for what he is and attempt valiantly to save what's left of GT football.