“I've found that prayers work best when you have big players.”—Knute Rockne
Successful college football recruiting is simply attracting the most skilled, talented, and biggest players to your team. The Arms Race for these high school athletes each year attracts more exposure on Internet sites, creates more recruiting experts, and consumes more time for fans, coaches and athletes. Coaches realize how much recruiting can impact their programs no matter what the sport.
"I'd rather have a lot of talent and a little experience than a lot of experience and a little talent.”—John Wooden
Can’t we just predict which recruit to pursue? Statistics to the rescue. Three football-junkie economists from Mercer College developed a model using statistical analysis software for predicting the top recruits’ choices with uncanny accuracy. (2009 link) In 2005, using Rivals' top 100, they correctly predicted 68 percent of recruits' choices. Recruits averaged four schools. Not too shabby.
The pressure and tension in getting 17- or 18-year-old high school stars to commit to your university can become unbearable. Paul Bryant described the Recruiting Burnout that affects many coaches best: “I can remember loving to recruit. I knew I was going to do my best. But traveling and recruiting doesn’t appeal to me any more. It’s not as much fun as it used to be.”
By tweaking what factors do matter, the Mercer three have increased their success rate to 73.2 percent, expanding it last year to the Rivals 250, with recruits averaging 6.2 schools! Dude! Almost three out of every four players. For a statistician, that makes recruiting more fun each year. Even the Bear would have loved those results, though maybe not the increasing recruiting time each year.
What factors influence recruits' decisions? This generation’s Rocknes, Woodens, and Bryants are listening. Those factors turn out to be predictable, insightful, and surprising.
A few factors are almost predictable:
— Whether the school is in a BCS conference
— Whether the recruit made an official visit to the college
— The number of conference titles the school recorded in recent years
— The final AP ranking of the school—but only in the previous year. Recruits seem to have short memories.
— Whether the school is in the same state as the recruit
— The distance the school is from their hometown
Universities in football-rich states—California, Texas, Ohio, Florida—with numerous Rivals 250 prospects should be favored. The first rule in recruiting should be, “We will protect this house.”
Some factors have to do with facilities:
— The size of the college’s stadium in seating capacity
— Whether the school has an on-campus stadium
— The current age of the stadium
The ideal recruiting situation for a coach would be to represent a BCS school, in a talent-rich state, either in the same state and near to the recruit’s home, ranked high in the previous year’s AP poll, with large, modern on-campus facilities, with highly-ranked athletes recruited at the prospect’s position the previous year. This is a tendency of the rich getting richer.
A recent study by Sports Illustrated found that, “Of the nine schools that won 50 or more games from 2004-08, seven signed more than half their recruits during that span from within their state or from within 200 miles of campus.”
Texas rarely recruits out of state—93.2 percent of their roster are in-state. Only seven non-Texas residents are on UT’s current roster. The Longhorns are the big dogs of Texas, a state where the top 48 recruits in 2009 carry a four or five-star rating. “Bag ‘em and Tag ‘em” is the slogan heard in Austin in June, as recruiting the best in Texas winds down.
(Rival’s 2008 and current 2009 recruiting rankings in parentheses)
Those football-rich programs must also fight off neighboring raiders—Oklahoma (6, 11), LSU (11, 1), and Georgia (7, 7). Oklahoma has 72 non-residents on the football roster with 42 from Texas, while 32 are from Oklahoma. The annual Red River Shootout impacts more than the conference championship, but recruiting as well. Oklahoma stole the top three Texas recruits in the 2008 class and has three of the top 14 Texas ’09 recruits.
LSU has 38 non-residents on their roster, with 15 from Texas. LSU has verbal commitments from two of the top four Texas recruits this year.
Florida (3, 10), Miami (5, 9), and Florida State (9, 13) carve up Florida athletes based on state geography. (Rival’s 2008 recruiting rankings in parentheses) Florida, Florida State, and Miami have 33, 32, and 23 non-resident football players, but only seven, six, and eight reside outside the South and Texas. Georgia has eight Florida residents on their roster and has gotten the commitment of the top quarterback from Florida in 2009.
Ohio State (4, 2), USC (8, 5), and Texas (14, 3) dominate their respective states. For 2009, Ohio State already has seven of the top nine committed players in Ohio. USC has the top five California players. UCLA (13) scoops up more of the California recruits. The Bruins have only 12 non-California residents on its roster. USC has 21 non-California residents.
Here's a breakdown by Rivals of who is winning in the top states in 2009.
The following factors, to the surprise of the Mercer economists, “don’t systematically influence the decisions of high school athletes”: the number of Bowl Championship (BCS) appearances, the number of players from the college drafted by the NFL, the number of national championships won by a school, a school’s graduation rate, and, curiously, the current roster depth at the player’s position.
The economists found that highly-sought recruits were statistically attracted to schools that had recruited one or more top recruits the previous year. Many recruits say they want early playing time, but then choose successful football programs loaded at their position. That may explain why Southern California can recruit some of the nation’s top running backs each year.
Which football programs must work the hardest and have shown the greatest success? Ones without conference titles, far away from a recruit’s hometown, in less football-rich states, and who were not ranked highly the previous year—even if they have numerous National Championships, BCS appearances, a number of players drafted by the NFL, and a high graduation rate.
Looks like Notre Dame and Alabama would be in real trouble in 2008. Far from it. Both achieved recruiting success against statistical odds.
Notre Dame (1), Alabama (2), Clemson (12), and Colorado (15) finished out the top 15 in 2008. Only Clemson was ranked in the 2007 AP’s top 15. Notre Dame, in 2007, had their worst season ever but their best recruiting class in history. The '08 class of 23 players averaged almost four stars (3.96) out of five—tops in the nation. Alabama signed 32 players (average stars: 3.72), but have ended up with a class of 27 due to attrition.
What’s brewing in these programs? How are recruits attracted to Indiana, Alabama, South Carolina, and Colorado?
All have been rebuilding with coaching staffs with three years or less at these institutions. Alabama and Notre Dame are rebuilding on tradition-rich and nationally known football programs. Nick Saban and Charlie Weis both surrounded themselves with great recruiters. Clemson’s Dabo Swinney possesses an excellent recruiting and coaching staff. Dan Hawkins has brought optimism, competition and pride back at Boulder.
A detailed look at the Mercer’s Prediction Breakdown in 2008 shows these same football programs stole recruits from other programs based on predictions, bucking recruiting trends. While the statisticians correctly predicted 73 percent of recruits’ choices, the top football programs who unpredictably took the remaining 27 percent of prospects were:
— Miami: won the most predicted prospects from other schools with eight, taking three from Florida
— Alabama: won seven, including four from Auburn
— Clemson: won six, including two from South Carolina
— Colorado: won four
— Notre Dame: won three
But almost all have to fight the distance factor. Of the five schools, Notre Dame and Colorado fight the distance factor the most.
— Notre Dame: 20.8 percent from within 200 miles, 5.9 percent from in-state, average distance—640.9 miles
— Colorado: 21.1 percent, 21.1 percent, 888.99 miles
— Clemson: 48.2 percent, 35.17 percent, 244.4 miles
— North Carolina: 56.8 percent, 44.1 percent, 301.8 miles
— Alabama: 70.2 percent, 54.2 percent, 211.9 miles
Their model this year, in the case of the nation’s top inside linebacker, Manti Te’o from Hawaii, goes against the distance factor. Te’o is predicted to go to Notre Dame (32.1 percent) over USC (22.1 percent), Stanford (18.8 percent), and UCLA (17.5 percent).
The South’s high school football talent is indeed rising again and benefits their recruiting rankings. (map link) Nine of this year’s Rivals' top 15 teams in recruiting are from the Deep South. Including Texas and Oklahoma as part of the South, only four non-South teams—USC, Ohio State, Michigan, and Michigan State—crack the 2009 top 15.
How are the Mercer trio doing this year in predicting recruits' choices for the Rivals 250? 76 percent—with one week and 37 high school prospects to go. Is that because Notre Dame, Clemson, and Colorado have dipped in the rankings? We’ll know for sure in another week when the letters of intent are all in.
Of the three schools' commitments in the Rivals 250, only one recruit’s choice (for Notre Dame) was not predicted—compared to 13 last year.
Until then, it’s nice to know that statisticians can’t always predict all behavior. Some prospects may choose schools based on future success outside of football, graduation rates or even national championships. That won’t stop the Mercer trio won’t stop trying to even more accurately predict the choices of these talented teenagers—something every coach would like to do.