|
Articles
|
Recruiting rankings and success on the gridiron Richard Linde, 10 February 2006 Not only can the avid college football fan feed on a steady diet of televised football games during the season, but he can follow his favorite team's progress in corralling its next generation of talent by following the recruiting services on the Internet. These services, such as scout.com, rivals.com, and realdawg.com, provide all kinds of data relating to student athletes at the high school and junior college levels. One set of data, the recruiting services provide, compare each school in the country with the others, showing their relative ranking in signing student athletes to letters of intent for their next football classes. To provide these rankings, each prospective student athlete is ranked qualitatively (given one to five stars) and quantitatively (ranked numerically by position). The Internet and recruiting Presumably, if 15 of the top 25 schools in the nation from a past season have offered a PSA a scholarship, he would be given 5 stars and receive a high positional ranking, regardless of whether a scout from a recruiting service had seen him or not. The ratings of would-be collegians are stored in centralized data bases on the Internet, which are available to everyone, even college coaches and the athletes themselves. A PSA's stock may rise as more coaches show an interest in him. Many recruiting sites provide highlight videos of each PSA, along with a list of the schools that have shown an interest in him or have offered a scholarship. Sometimes a PSA may orally commit to a school, then change his mind and sign with another school. Usually, though, PSA's honor their oral commitments. But is a successive series of recruiting rankings for a team, provided by the recruiting services, the most important factor for predicting its success on the gridiron? Obviously, quality recruits make for a quality team. However, the data below, taken from recruiting data made available from scout.com, suggest that other significant factors may influence a team's success as well. The correlation between rankings and success on the field. Winning battles in the recruiting wars, however, are important determiners for future glory. For example, over the 2002 through 2005 recruiting seasons, Southern California had an average national ranking of 5.5, according to data obtained from scout.com. In that period, it won two national championships and lost by a phantom touchdown to Texas in the BCS title game this season. The effect of USC's last several recruiting classes correlate significantly with its success on the playing field (see Table 1. below). National champion Texas averaged 9.5 over the four-year period. Oklahoma which played in the previous title game averaged 4.25 ((2+3+7+5)/4). That is, the Sooners finished second in the nation in recruiting in 2002, third in 2003, seventh in 2004 and fifth in 2005 for an average of 4.5. Ohio State (11.5) beat Notre Dame (18.5) in this season's Fiesta Bowl. In the Orange Bowl, Penn State (26.5) upset perennial powerhouse Florida State (6.25) in overtime. As a result, Joe Paterno brought in a class ranked sixth in the country. Over the last five years, Coach Pete Carroll of USC has brought in 43 student athletes that belonged to the nation's top 500, according to scout.com. Carroll corralled 12 of the nation's top 100 PSA's this season alone. None of Carroll's players who are headed for the NFL draft this season would have been considered sleepers when they were recruited. All of them were highly sought after, with 4-and-5-star ratings. The data from the two tables below suggest that recruiting rankings are becoming more of a determiner for predicting success on the field than in the past. For one thing, the combines are separating the wheat from the chaff, functioning like the SAT, which is a predictor for success in the classroom. Thanks to a lackluster performance in the combines, an all-everything quarterback from Riverside County failed to get a D1 scholarship this year. Conversely, an overlooked PSA can make himself known at a combine. The combines and training camps provide accurate data relating to a prospective student athlete, e.g., height, weight, 40-yard dash time, bench press stats, vertical jump ability, and 20-yard shuttle. Participants are separated into position groups to work individually on football skills and techniques with top-notch position coaches. The data accumulated are available on most recruiting sites. The five teams in the Pac-10 with winning records last season (USC, UCLA, Oregon, California and ASU), according to data taken over the 2002-2005 recruiting seasons, have an average national ranking of 24.3. The teams with losing records (Stanford, OSU, UA, WSU, and UW) have a national average of 38.8 over the same four seasons. Seven of the top 10 teams in the final BCS poll, shown by Table 2 below, averaged 11.3 for the 4-year recruiting period (2002-2005), each team's 4-year average being in the top 20. Penn State (26.5), Oregon (30.0), and Virginia Tech (27.5) were anomalies. Other factors affecting won/loss records Other factors affecting a team's won/loss record are strength of schedule, home-field advantage, coaching stability, injuries, early attrition, luck, and NCAA probationary periods and sanctions. For example, Oregon (10-2) apparently overachieved last season, considering its four-year average of 30.0. However, it has a distinct home-field advantage playing at Autzen Stadium, and boasts of having the dean of Pac-10 coaches, MIke Bellotti. Last season, Bellotti installed a spread offense, which quarterback Kellen Clemens ran nearly to perfection before getting hurt. Having playmakers might explain why Oregon has done so well despite its bland recruiting average. In 2002, Oregon recruited DT Haloti Ngata, a five-star recruit, who was rated number one in the country at his position by scout.com. As a freshman at Oregon, Ngata started in 12 of 13 games. Oregon led the Pac-10 in defense in the 2005 season. A big part of that defense, Ngata was the twelfth overall selection in the first round of the 2006 draft, when he was taken by the Baltimore Ravens. Oregon signed 5-star RB Jonathan Stewart to its 2005 class; scout.com ranked Stewart number 3 in the country at his position. Stewart qualifies as a super-playmaker, having impressive numbers in the 40, weight-lifting categories and vertical leap. Over that same four-year period, Washington, which had the third best recruiting ranking in the Pac-10, finishing behind USC and UCLA, had a 16-31 record to show for its recruiting efforts. The coaching carousel at Washington -- three different head coaches in the four-year segment -- partially explain its record. Although Washington's 2004 class was ranked 22nd in the nation, three of its 4-star members never enrolled in school. During the regular season, West Virginia (11-1), a member of the Big East Conference, beat 9 Division IA teams that had a combined record of 47-55. It also posted a 35-7 win over Wofferd, a non-Division IA school. Its narrow victory over Georgia (7.5) in the Sugar Bowl qualifies as a huge upset, considering the Mountaineers' mediocre recruiting number (44) and its comparatively weak schedule. The Mountaineers finished 11th in the BCS Poll. From a correlative standpoint -- that is, comparing the winning teams in the Pac-10 with the losing teams -- this is as good as it gets, for recruiting rankings are not the whole enchilada when it comes to predicting success on the playing field. The data presented in Table 2 seems more meaningful from a correlative standpoint than the Pac-10 data, the 3 BCS anomalies notwithstanding. Author's note: Comparing a successive series of recruiting rankings with a specific team's won/loss record for one season is much less meaningful than if a contiguous series of rankings were compared with its won/loss record over a period of years. Too many of the other variables listed above can influence an outcome in a single season. In other words, collegiate teams don't play enough games in a single season to make valid statistical comparisons, in my opinion. These comparisons will become more meaningful as more data are collected in subsequent years. The results of future analyses might even factor into the BCS formula someday if the mathematical gurus want to take another shot at strength of schedule. As for me, I’m not statistically qualified to make any "significant" evaluations, other than the simple comparisons I made.---------------- In Table 1 below, Pac-10 teams are ordered by their won/loss record for the 2005 season. The column depicting the four-year average consists of the 2002, 2003, 2004, and 2005 seasons, while the last column is for all 5 years tabulated. The other columns show the team's recruiting ranking for the year. For example, Washington had the 35th best recruiting class in the nation this year. It will be interesting to see how the last column (the 5-year average) will correlate with the won/loss records of conference teams in 2006. According to that data, USC, UCLA, and Cal will most likely vie for the conference title. The data do not show that USC must be rebuild an arsenal decimated by early attrition, which is why we armchair prognosticator's must intervene. The data also suggest, based on its last two classes, that Arizona may be among the top five conference teams next season. Table 1. Pac-10 teams and their national recruiting rankings over the last five years. The pseudo-nicknames for the schools reflect the hard feelings that exist among their fans. Guess which school I attended?
Table 2. Final BCS Rankings and Recruiting Rankings
Richard Linde (a.k.a., Malamute) can be reached at malamute@4malamute.com |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
