Recently I was listening to the Solid Verbal podcast and the hosts got into a brief discussion of conference strength. This happened during a conference preview show and while they didn’t spend too much time on the subject (they were more focused on specific team by team analysis) the time that was spent on it seemed to be a little too “gut level / touchy-feely” for my tastes. Specifically, Dan Rubenstein said that he felt the ACC was the strongest conference this year based on the fact that it has multiple “Top Tier” teams (Clemson, Miami, FSU) a deep middle class, and a lack of really bad bottom feeders (here’s looking at you Rutgers and Illinois). In general I don’t really have a major dispute with the points of this argument, but I do question how those points equate to strongest conference. So it got me thinking: why isn’t there some type of objective/stats-based conference ranking system? Since we have metrics like FEI and S&P+ that rank all 130 teams you’d think it would be pretty easy to take those numbers and determine a conference ranking. Well, either my googling skills are pretty bad - in which case please feel free to skip the rest of this and just drop down to the comments and blast away - or no one has taken the time to do the work. So, here is my attempt to do some work.
As I pondered the best way to turn either FEI or S&P+ into a conference ranking one of the first things that came to mind was utilizing a cross-country style scoring system. For those not familiar, a cross country meet is scored by giving the 1st place finisher 1 point, the 2nd place finisher 2 points, 3rd place 3 three points, etc. Each team’s 5 fastest runners are scored and the lowest overall score wins the meet. One issue in using this type of system is deciding how many teams would be included in a conference’s score. Using only the top 5 (a la cross country) would tell you which conference is strongest at the top but I don’t think many people would argue that is the best way to measure conferences. Using all teams in a conference is problematic due to the differing sizes of conferences. The other drawback of the cross country style system is that “margin of victory” is not captured (i.e. being 30 seconds faster than the guy behind you counts the same as being 1 second faster). As I was pondering this second fact the obvious hit me – why not just use the damn numbers? So that is what I did.
I decided to use S&P+ as the basis for the ranking, simply because it’s available in a 2018 pre-season form. Not sure on the schedule for updating FEI, but as of now it is only available for the 2017 Season and before. Once FEI ratings come out I will probably do something similar for that system as well. Ranking the conferences by S&P+ was a simple matter of calculating the average of all the teams – I think??? I’m no Nate Silver Ed Feng so perhaps I’m making it too simple, but given that a team’s S&P+ rating effectively measures how many point better they are than an average team, averaging this “points total” over a group of teams seems to be pretty sound. Once again, feel free to correct me if I’m wrong. So what do the numbers say?
Table 1 – 2018 Preseason S&P+ By Conference
Takeaways / Observations / Random Musings
- The SEC comes out ahead pretty clearly. Shouldn’t be a big surprise, though there’s probably a few of us that thought the Big Ten would be higher. One through five (OSU through Wisconsin) the Big Ten is actually better than the SEC (20.3 vs. 20.2), but after that . . . not so much. Interesting that after the SEC the other four conferences are pretty closely grouped. I think this is an excellent example of why a ranked list of five teams is so silly.
- Similar story in the divisional battle of Big Ten East vs. SEC West, though those two divisions are clearly the cream of the crop. The ACC Atlantic isn’t too far behind. All the other divisions are well behind those three. Big Ten West??? Woof.
- My biggest take away from looking at this? UCF needs to STFU. Holy HELL is there a gap between the power 5 and the group of 5. If ever you were looking for a reason to create a separate division of football for the top 60+ teams this is it. We have two distinct groups here with a clear delineation between them.