Normalized Probabilities Case 3, Sub-analysis 1: 2006 - 2009 Rivals 250 Lists as Population Source, Scaling Competition by Conference

Just as in the "normal" Case 3-1, this study scales schools by their conference affiliation. Given that it follows the normalized probability model, the results are somewhat different than the effective population Case 3-1. As discussed in the methods section, the maps are not very interesting, it is the tables rather that have the real interesting data hence they are shown immediately below with the maps below them.

The population source--the Rivals 250 lists--makes this particularly interesting because we know where these recruits actually chose to enroll and we can assume that every program wanted them. Therefore the numbers calculated--which represent how many of these Rivals 250 recruits a given school should have attracted given the only two criteria are geography and conference affiliation--can be compared to what actually happened.

It's important to first compare the results to those from the effective population version of this case. For the most part, the results are not significantly changed for teams in the Big 6 conferences. The normalized probability method takes into account the competition from other schools in a similar fashion (but not exactly, hence there are subtle differences). For the non-AQ schools, however, the results are quite different as their recruiting influence is scaled down. That is because the effective populations method quantifies how good of a recruiting location they could have whereas this method actually quantifies how many recruits they are liable to attract. So when their recruiting influence is scaled down to zero, this probability model suggests they will get zero recruits (with the exception of Notre Dame, this is pretty much true for Rivals 250 recruits).

Now let's look at the results. As mentioned, non-Notre Dame schools do not attract many of the top-notch recruits. For this reason it's best to isolate the Big 6 teams and see how they fare in the model when non-Big 6 teams are scaled down to zero versus how they fared in the real world. The most obvious underperformer is Georgia Tech which ranks as the easiest place to recruit in this model but somehowe finished behind 33 other teams in attracting top recruits. They did do okay, however, and in terms of net recruits where they are only the 14th biggest underperformer. It's easy to characterize the overperformers. The top 5 reads like the top 5 teams of the decade in terms of on-field performance: USC, Florida, Texas, LSU, Ohio State. These top few teams exceed their expected number of recruits by huge margins and as a result there are more underachievers than overachievers.

It's interesting to compare the over/underperformer lists (Table I) to see where the model goes wrong--or what other variables should improve it. Things that may cause you to underperform in recruiting:

• Being a private school in the south (Vandy, Wake, Baylor, Duke)
• Being a "basketball school" (Duke, Indiana, Kentucky, Louisville, Kansas)
• Being overshadowed by another state school (Texas Tech, South Florida, Cincinnati, Iowa State)
• Having someone not consider Notre Dame a Big 6 school (Indiana, Purdue) -- I'll have to repeat this with Notre Dame considered as a Big 6 team

On the flip side, what helps recruiting? Winning recently and having won in the past. The Top 12 overperformers fit that bill pretty well. In sum, recent winning percentage and total number of wins in team history may be two important scaling factors to add in (the bottom right of Table II actually shows a calculated error for this analysis).

Now let's look more at the actual numbers and what they say about recruiting in the country as a whole (Table IV). The top schools are basically in the heart of the SEC--Florida west to Mississippi (Ole Miss and Mississippi State are notable underachievers in recruiting). It's more interesting to see how far down the list some prominent programs are (note that this list really only goes to 65, the number of Big 6 teams). Even Texas only finishes a pedestrian 33rd when non-Big 6 schools are not considered (scale factor = 0). UT fans will at least be happy to note they are one spot ahead of Oklahoma. Nebraska, one of the two dominent teams of the 1990's, is a mere 54th out of 65. I'm a bit surprised the west didn't catch up more in this analysis as I expected it to correct for low population density states. Again the Oregon/Washington foursome bring up the bottom of the list. I hope the Oregon schools are paying their coaches well.

A note about the images. The circles indicate the number of recruits and location of the 120 schools. They're pretty straightforward except for when non-AQ schools are scaled to zero, then the images don't distinguish "good red" from "bad red". Keeping in mind that the top schools are all in the southeast should help you make sense of it, however.

Table I - Under and Overperformers

Table II - Actual vs. Modeled Number of Recruits, Teams Sorted Alphabetically

Table III - Recruiting Ranking, Actual vs. Modeled, Teams Sorted Alphabetically

Table IV - Rankings for Each Scale Factor