A Tech Effect?

Technical Recruiting Analysis Suggests Technical Schools Bad at Recruiting

It’d be one thing if their on-field performance lived up to the stereotype of the athletically-challenged engineer but that’s simply not the case. So then, what, if anything, explains the fact that Georgia Tech, Texas Tech, and Virginia Tech compose three of the six worst-recruiting schools in the country?

Of course I must justify my throwing down the "bad recruiter" gauntlet and explain how I arrived at a quantification of schools’ recruiting success. There are many ways of going about this, but the method I chose is to look at the factors that drive recruits (in this case: distance between recruit and school, being in a BCS conference, teams’ winning percentages (over previous 3 seasons), prestige rating, being in the same state, being a private school (which tend to recruit nationally; also: a nice Notre Dame/USC fudge factor)) and see if you can predict how many recruits will choose a given school. The relative weights of each of those variables can be optimized by testing with a large dataset (in this case: the Rivals 250 lists from 2006-2011). End result: probabilities of every recruit attending each school; sum them up and you have a forecasted recruiting haul over the same 2006-2011 period for every college.

And then you have the actual recruiting haul by all of those schools to compare to the model. Some schools do better, some do worse—and so I arrive at my harsh judgment: the three Techs (sorry Louisiana Tech, I left you out of this one) compose three of the bottom six slots for underperformers, with Georgia Tech last of all.

While their recruiting has been subpar, their actual football playing has been pretty solid, and to the extent that a history of winning is a strong predictor of future recruiting, these schools are victims of their own on-field success. Until last season, Georgia Tech hadn’t had a losing season since 1996. Virginia Tech and Texas Tech haven’t had losing seasons since 1992, with the former having been consistently elite in that timeframe. Although GT and TT both took steps back last year, as shown in the table below, in the time frame of the analysis all three schools did pretty well by the metric of their ranking from Sagarin’s predictor algorithm (which rather than mere winning percentage includes strength of schedule and margin of victory/defeat).

To cut to the chase, my model forecasts that 35 (out of 120) teams will bring in 12 or more Rivals 250 recruits (or 2 per year on average from 2006-2011). Of these 35, only 6 schools failed to yield 50% as many recruits as forecasted—the three Techs, West Virginia (a big-time underrecruiter), Missouri, and Wisconsin (see table below). The bottom three in terms of differential between forecasted and actual results are GT, West Virginia, and VT (in that order).

The three Tech schools have something further in common: unlike the other three schools they are in football-heavy states and in good proximity to other recruit-heavy locales. This is somewhat less true for TT and VT which reside on the western and northern edges, respectively, of the Texas-ACC/SEC corridor than for GT, which is located in the corridor’s epicenter, but regardless these three teams are located in top-10 states for producing top recruits, as shown in the table below. Note that in comparison the states of West Virginia, Missouri, and Wisconsin rank tied-41st, 22nd, and tied-30th, respectively.

To sum up: the three Techs win often and consistently, reside in rich recruiting grounds, but fail to bring in their share of top-level recruits. It’s important to point out that this "failure" could also be a symptom not of being turned down by recruits but by either consistently overperforming or better evaluating prospects (and hence not wasting resources going after a few high-profile recruits), but I don’t think this fully explains the gaps.

A naïve interpretation would be that these "tech" schools push recruits away for academic reasons. While VT and GT, in particular, are nationally-regarded engineering schools, that is not all they do, nor is it the case that they only going after highly-motivated science-oriented student-athletes. According to the most recent NCAA academic progress rate statistics (APR), VT received a mediocre 940 and GT a better-than-most (but not in the upper echelon) 967. To corroborate my analysis of VT’s student-athlete threshold, I will merely say "Marcus Vick". Texas Tech, which doesn’t quite have the national reputation of GT or VT, received an APR score of 944 (a point out of 6th place in the Big 12 last year).

So it’s not any sort of academic obstacle to pursuing all the recruits at these schools’ disposal. What about other connotations associated with tech schools? Perhaps the lack of females? Maybe we have something here. The table below shows the male:female ratio at the three Tech schools versus their respective conference average (note: I included Colorado and Nebraska in the Big 12 for this).

Does the lack of women at these schools dissuade recruits? Even the most skilled recruiting host can’t disguise such a dearth of females during an on-campus visit. Well, let’s be honest, I doubt this tells the whole story, but I bet it isn’t irrelevant either. More than simply a gender numbers issue, I speculate that these high ratios are both caused by and contribute to a certain campus culture. I wonder if high-profile recruits are less likely to feel like a BMOC when they visit the Tech schools due to this culture. But I’ll leave any further analysis on this front to those with more sociological chops than myself. These schools definitely stand out as outliers; I’d love more insight on these schools’ recruiting philosophies and successes from some fans who know these schools well.

I hate to knock schools that have done a pretty good job of winning. So I’ll close with a final stat that demonstrates the superior importance of coaching over recruiting rankings—average win percentage over the same time frame as the 2006-2010 recruiting classes for the six ACC schools with the largest number of high-profile recruits.

Edit: I meant to reference Purdue as another school that while not "tech" by name is similarly engineering and male-heavy. A commenter pointed out Texas A&M and Oklahoma (A&M) State as well. Purdue fits very well into the category of the Tech schools mentioned (under-recruits, high male:female ratio) while A&M and Oklahoma State are a mixed bag--neither is particularly male heavy, and OK St. actually recruits quite well. A&M is a big underperformer but they seem to be on the upswing. If anyone is aware of any other male & engineering heavy I-A schools, let me know.