clock menu more-arrow no yes mobile

Filed under:

The Monday Mathematical (Wednesday Edition): Roster Construction

No models required, just visualizing the rates at which recruits catch on

NCAA Football: Kentucky at Tennessee
Crouched low to make a sneakier getaway out of Tuscaloosa
Randy Sartin-USA TODAY Sports

As we discussed in the wake of Jalen Hurd’s decision to transfer, 12 players from the 2014 recruiting class have left early. That got me thinking...what does the structure of Tennessee’s roster look like? Which classes make up what percentage of the team, and how does that compare to other schools around the league?

Per the subtitle, there’s not really any math here other than looking at simple ratios, but I wanted to do two things. One, I wanted to get a sense of whether or not the transfers and turnover have been unusual compared to our peers. Two, I’ve been messing around with Tableau at work for visualization purposes, so I could use the practice! I suppose I should also mention that this isn’t a sophisticated recruiting analysis...I’ll leave that to Hunter and the experts.


I chose to look at three teams: Alabama, Georgia, and Tennessee.

The choice was less about anything related to the success of Alabama and Georgia and more about the tenure of their coaches. In theory, a current football roster could be made up of individuals from the recruiting classes of 2012, 2013, 2014, 2015, or 2016. Nick Saban is one of only two coaches in the league who have more than 5 seasons at their school (Dan Mullen); thus, he is solely responsible for all of the classes that could currently contribute to Alabama’s roster. Given that Butch came on the scene early in that stretch, I wanted to balance him with somebody who came on the scene late in that stretch in Kirby Smart.

From there, it’s just a matter of using the ol’ rosters and the ol’ Google to figure out who is on a team, who graduated, who transferred, who was dismissed, and who went on to the league.

For the purposes of this article, players are categorized by what caused them to leave school...if they ran out of eligibility, I’ve noted them as “graduated” regardless of whether they were subsequently drafted. Only players who left early for the draft are noted as “NFL.”

Breakdown of Vols Recruits

From 2012 to 2016, Dooley and Butch combined to recruit 130 players. Of those, 75 players are currently on the roster, 31 transferred, and 24 others departed amidst other circumstances (dismissed, did not academically qualify, declared medically ineligible, left early for the NFL, ran out of eligibility, or left the team for personal reasons).

Waterfall chart of outcome of UT’s 2012-2016 recruits

At first glance, transfers look a bit high. A full 24% of recruits brought in since 2012, 31 out of 130, have left the program. Thankfully, I’ve got just the thing to put all of these figures in context!

Comparing the outcomes of 2012-2016 recruits across Bama, UGA, and UT

The colors on this chart represent the different recruiting classes, and then each set of clustered columns explores how many recruits of the three teams in question are classified in each bucket. For example, of the 75 recruits still on UT’s roster, 3 are from the class of 2012, 10 are from the class of 2013, etc. (Note: this would be much more engaging in an actual Tableau workbook, but c’est la vie.) 78 players from Bama’s classes are still on the roster and 77 from UGA are still on the roster.

From a pure outcome standpoint, Tennessee’s classes have fallen out about like Alabama and Georgia everywhere except transfers. 17 players transferring out from the classes of 2012 and 2013 could be explained by the Dooley-Jones coaching transition, but UGA has not been similarly impacted thus far from the Richt-Smart change (that may be worth keeping an eye on).

What’s more concerning is the nine transfers from the class of 2014. Six of those nine were 4*, blue chip recruits. You can survive those kinds of defections if they’re replaced with more star power, but thus far, that has been an issue for the class of 2017. When you lose highly ranked players and struggle to replace them, you end up with...

Jimmies and Joes, man...Jimmies and Joes.

Of those 75-80 recruits in the last 5 classes that have stuck at each school, UT has by far the smallest percentage of 5* (1.33%, which represents one player [Khalil], while UGA has 10.4% and Bama has 21.8%) and the smallest percentage of blue chips (48% for UT versus 55.8% for UGA and an obscene 78.2% for Bama). That’s...going to be a problem for long-term success.


Firstly, I’ve learned that Tableau is counter-intuitive.

More importantly, we’ve seen that the volume of transfers under Butch has been higher than that under some of his peers in the SEC. Will that ultimately translate to anything?

Maybe...sometimes you have to let these things play out, but this visualization helped me contextualize what we’re talking about. As we’ve seen this year, having a deep stable of talented players is crucial to being able to withstand the rigors of a football season. Having them leave your program at a high rate won’t help that.