Crunching numbers to determine if top 100 recruits lead to BCS bowls

Taylor Mays and the other top recruits at USC are slumming in the Emerald Bowl this year. Icon SMI

The Count feature in the Wall Street Journal typically offers a fascinating look inside the numbers of sport. If you've read Freakonomics, you understand the concept; use raw data and sprinkle in a little counterintuitive thinking to draw eye-opening conclusions.

So when Rupert Murdoch's most legitimate baby dipped its toe into the murky abyss of football recruiting rankings Tuesday, I was understandably excited. I clicked, hoping to be enlightened. The story's findings were indeed eye-opening. Author David Biderman analyzed this year's 1,496 bowl starters and their teams' bowl destinations and concluded that recruiting highly ranked players doesn't necessarily help teams reach BCS bowls. Unfortunately, the fact Biderman pegged the story on -- that USC, with a multitude of starters who were ranked in the Rivals.com top 100, is playing in the Emerald Bowl against a Boston College team that has no starters who were ranked in the Rivals100 -- turned a blind eye to some critical data. Biderman never mentioned that USC, with players who are still on this year's roster, won the Pac-10 and played in a BCS bowl each of the past three years.

Biderman didn't do anything wrong other than use too small a sample size. Still, his story inspired me to do some research of my own. Curious as to just how accurate the Rivals100 has been, I looked up each player ranked in the top 100 from 2006 to 2009. Expanding on Biderman's idea of counting the players who matured into starters, I catalogued the contribution each made to his team in 2009. I also catalogued whether the team that signed the player had played in a BCS bowl in the past few years.

Are the findings conclusive? Probably not. As Biderman pointed out, neither Boise State nor TCU starts a single player ranked in the top 100, and most poll voters consider those programs to be two of the best five in the country this season. That has more to do with how Rivals rankings work than the quality of the players on either team. Rivals analysts make firsthand evaluations of players, but they also consider which college coaches chase each particular player. If Pete Carroll, Bob Stoops, Nick Saban and Urban Meyer -- coaches with proven track records -- all want the same player, then that player must deserve a four- or five-star ranking. If Boise State and Colorado State are the kid's only offers, then he's probably a three-star unless an analyst sees the player do something amazing with his own eyes.

Some of the rankings get goosed for commercial reasons, too. Though it's unlikely anyone at Rivals would say this on the record, if Virginia Tech fans bought Web site subscriptions at the same rate as Notre Dame, Alabama or Florida State fans, then the Hokies' recruiting classes might not be so underrated relative to their success on the field.

Despite all that, an analysis of the Rivals100 for the past four years proves the site's analysts are just as good as most college coaches or NFL general managers at projecting which players will shine after moving up a level. The data also show that if a program can accrue top-ranked recruits, it may not have a better chance of playing in a BCS bowl this particular year, but it stands a far better chance of playing in BCS bowls on a consistent basis.

Of the 400 players ranked in the Rivals100 since 2006, 101 signed with a team that will play in a BCS bowl this year. Meanwhile, 179 (44.8 percent) signed with a team that has earned at least two BCS bowl berths (USC, Oklahoma, LSU, Florida, Virginia Tech, Texas, Ohio State, Alabama) in the past four seasons. During that stretch, the top four stockpilers of Rivals100 talent (USC, Texas, Florida and Notre Dame) have combined to earn nine BCS bowl berths.

Obviously, just signing with a bunch of other highly ranked recruits doesn't guarantee a player will find individual success. Florida leads the way with 12 Rivals100 starters, but Texas and Alabama reached the BCS title game with six and five starters, respectively. The Journal uses different numbers for starters, but that's understandable. Starter is often a relative term. So instead of just cataloguing starters and non-starters, I broke down the signees' 2009 contribution into five categories.

Starter: Started most of his team's games, or started for much of the season before suffering an injury.

Major contributor: This could be a nickel back, a No. 3 receiver, a return specialist, a second-rotation defensive lineman or an offensive lineman who backs up three or four positions. I also included backup quarterbacks such as the Longhorns' Garrett Gilbert and the Gators' John Brantley. Any player who is a batch of bad shrimp away from pulling the trigger on one of the nation's best offenses is a major contributor. Tennessee tailback Bryce Brown is another example. The class of 2009's top-ranked recruit isn't the starter -- that would be Montario Hardesty -- but anyone who watched Tennessee play knows Brown contributed significantly to the Volunteers' improvement.

Contributor: This player may or may not have lettered, but he played occasionally -- often on special teams.

Non-contributor: This player did not play for the team with which he signed in 2009. He may have transferred, gotten injured or moved on to the NFL.

Redshirting: It's self-explanatory. And contrary to popular belief, the redshirt isn't dead. Thirty-eight of the top 100 in the class of 2009 redshirted this season.

Texas and Alabama may have only 11 Rivals100 starters between them, but they also have a combined nine major contributors and 13 contributors. In other words, 33 of the schools' 46 Rivals100 signees in the past four years made some sort of on-field contribution to help the teams reach the BCS title game.

Obviously, if a top 100 recruit wants to be a starter or a major contributor. That's tough at some programs, so transfers are inevitable. In the class of 2006, 17 Rivals100 players transferred to other four-year institutions. Another, former Maryland quarterback Jeremy Ricker, played for a professional indoor football team in his hometown of Harrisburg, Pa.

It would seem logical that USC, with 42 Rivals100 signees, would have the most attrition. That wasn't the case. Four of the 42 transferred to other four-year schools. One transferred to a junior college. Two were either dismissed or suspended for this season for academic reasons. The most attrition was at Florida State, where seven of 18 Rivals100 signees either never enrolled or left the program before completing their eligibility. Safety Myron Rolle has the best excuse; he was named a Rhodes Scholar in 2008 and is studying at Oxford this fall. Others either didn't qualify or left to seek more playing time elsewhere.

This might help explain FSU's slide. The Seminoles still have recruiting cachet because they finished in the top five every season between 1987 and 2000, but their on-field performance in recent years has shown that a player's ranking shouldn't simply be bumped up just because FSU is recruiting him.

The most curious case is Notre Dame (24 signees), which played in a BCS bowl in 2006 but has fallen off in recent years. If there ever was a program ripe to have its recruits overrated, it's Notre Dame. The Fighting Irish have a huge, passionate fan bases with plenty of members more than happy to fork over $9.95 a month to read a treasure trove of recruiting news. So it's good for business to keep the rankings high.

That still doesn't explain one fact. The main reason why Notre Dame recruits are ranked so high is that they also were pursued by other elite programs. So either the other schools' coaches evaluated the players incorrectly, or the players aren't reaching their maximum potential.

For the past two seasons, a chicken-egg argument raged as to whether Charlie Weis was coaching overhyped, poorly evaluated recruits or whether Weis had failed to develop accurately rated recruits. Notre Dame's 2009 team featured 11 starters ranked in the top 100 as recruits. With the exception of receiver Golden Tate -- ranked No. 101 in the class of 2007 -- the group included Notre Dame's best players. The players came in ranked higher than their teammates, and they turned out to be more talented than their teammates. They just couldn't beat other teams outside the closed system that was the Notre Dame practice field. That speaks to player development.

In other words, the data suggest athletic director Jack Swarbrick made the correct choice when he fired Weis.

Want to draw your own conclusions? Feel free to peruse the data for all 400 Rivals100 signees from 2006-09. Whenever possible, I tried to explain why a player left a particular program and where he went. Even if you don't notice any earth-shattering revelations, it's quite a trip down memory lane.