Well, it's been a while since I posted any statistical analyses, but here's one I've been curious about researching for a couple of years. With excitement ramping up for the upcoming college football season, I thought it was the right time to post this analysis of how the top conferences stack up against each other. If any new BTSC readers enjoy this post, you may want to review my RB Analyses (Part 1 and Part 2), WR Analysis, or QB Analysis.
For years now, I've been hearing the SEC touted as the best college football conference, particularly by fans/alum of SEC schools. Being from an ACC school and spending the early part of my life in northwestern PA, I'm not one to just agree that the SEC is axiomatically the best conference without question. So, I wanted to research the conferences to see how each one really stacks up against another. Now, before I get into this, I'm going to say that I really wanted to get more details particularly about each conferences' non-conference schedule, but for the most part I'm forced to rely on a couple of ratings systems that I have found on the web, plus a few other websites with articles and statistics about how the conferences rate against each other.
The majority of my analysis is based off of data that I got off of Jeff Sagarin's rating website and Teamrankings.com, a website created by Mike Greenfield, as well as little bit of data pulled from this Stewart Mandel article on this very topic.
My basic approach here was to determine the overall conference rankings based on Sagarin's rating system and the system used at Teamrankings.com, and see how dominant or balanced the conferences have been. I was able to gather data for the last ten years from Sagarin's site, but only for the last seven years from Teamrankings' site. I also reviewed the relative values assigned to each conference for the years to determine if there was a large or small relative disparity between the conference values. For example, one year the top conference may have a large margin of superiority over the next best, while in another year, the top conference's margin may be much slimmer.
So, let's start looking at the data from Sagarin's and Teamrankings sites and then we'll take a look at Stewart Mandel's article and data. First, we'll look at the Sagarin data:
Sagarin Conference Overall Rankings by Year
Sagarin Conference Rating Values by Year
Just a quick perusal of these tables indicates that the SEC has indeed lived up to its hype each of the last two years, finishing as the top conference each year, and also posting 2 of the 3 highest overall ratings values in the last ten years. However, the 8 years prior to 2006 indicate that the Conference relative strength has fluctuated between various different conferences. In fact, prior to 2006, three other conferences had finished at the top twice, which is twice as often as the SEC in that span. Also, the number of the ACC, Big 10, Big 12 and Pac-10 had finished in the top 3 was also right on par with the SEC in that span.
With that, let's take a look at the data from the Teamrankings website. First we'll look at the Overall Ratings just as with the Sagarin. Then, we'll take a look a conference Strength of Schedule and Non-League Rating (or rating versus their non-conference schedule).
Teamrankings.com Conference Overall Rankings by Year
Teamrankings.com Conference Ratings Values by Year
Looking at these tables, the SEC's argument appears even stronger, as Teamrankings.com ranked them the top conference for 4 out of the 7 years for which the data was available from the site. Unfortunately, we don't have the data from 1998-2000 to see if it also correlates with Sagarins' rankings. From looking over the data from both sources, there are some differences in the conference rankings, but, there are only 3 instances where the conference ranking differed by more than 2 between Sagarin and Teamrankings for a given year. There were 7 instances (out of a total 42) where the rankings differed by more than 1. So, with a couple of large discrepancies in 2002 and 2004, overall the rankings from both sets of data are consistent with each other. Based on this level of correlation, it seems safe to say that for at least two of the three years from 1998-2000, Teamrankings.com would have similar rankings as Sagarin. If that is the case, then we are again left to conclude that the SEC has been the dominant conference in the last two years, but prior to that, multiple conferences laid claim the title of the country's best football conference.
I didn't want to just look at overall rankings, because, I believe that what really determines relative strength between conferences is their non-conference strength of schedule and how they fair against that competition. Sagarin did not have that data broken out at all, and Teamrankings.com only had overall strength of schedule and non-conference Ratings. I was hoping to get to compare actuall non-conference strength of schedules, but I decided that this would have to do. So, here are the rankings and ratings values of each conferences overall strength of schedule and non-league (or non-conference) rating.
Teamrankings.com Conference Strength of Schedule Rankings by Year
Teamrankings.com Conference Strength of Schedule Values by Year
Teamrankings.com Conference Non-League Rankings by Year
Teamrankings.com Conference Non-League Ratings Values by Year
Looking at these numbers, what do they really tell us. While the SEC has the highest non-league rating 4 out of 7 years, it only has the strongest overall strength of schedule in one of those years (2nd in the other 3). Here is where it would be really nice to know exactly how the nonconference strength of schedules stack up. And that is where the data that I pulled from the Stewart Mandel article comes in. In his article, which I would recommend reading BTW, he looks at various different values to determine relative rank between the conferences. But, in my opinion, the best to measure each conference is through their non-conference results. So, from all of the measures that Mandel used, I pulled his Non-Conference RPI values for each conference and put it into a table format, shown below.
Nonconference RPI Values from Stewart Mandel article
Unfortunately, he did not break out his data year by year, so there we can't do a direct comparison, but this data is still worthwhile. According to this data, the Pac 10 has consisently, for the last 10 years played the toughest competition outside of their conference. For the years 1998-2003, the Big Ten and ACC both had better results against better competition than then SEC. Since then, both conferences have dropped in their non-conference achievements, being passed by the SEC and Big 12.
After looking at all of this data, what can we conclude with certainty? I would argue that the only thing we can really conclude is that the SEC has been the dominant conference for the last two years. However, prior to that, they were ranked 5th or 6th for the years 2004 and 2005 in both Sagarin and Teamrankings.com. And, going back to the 6 years before that, they were consisently a top 3 conference (certainly more consisently than any other conference). No conference really stood out as being dominant for any stretch between 1998 and 2005. According to Mandel's article, the Big 10 was the best conference from 1998-2003. But, according to the Sagarin data, the Big 12 did better, ranked in the top 2 for 4 out of the 5 years, while the Big 10 dipped as low as 5.
So, again, after looking at all of this data, the only thing really clear is that prior to 2006, conference strengths were pretty cyclical. In fact, the only conference that had very little change in their ranking was the Big East, who finished in last place every year until 2006. In the last two years, the SEC as clearly dominated, moreso than any other conference at any point in the past 10 years. And, that is probably to be expected when, according to Mandel's article their coaches' salaries have risen an average of 75% in the last five years. The average salary in the SEC is now $1.85 million, with 3 coaches making more than $3 million/year. So, for the foreseeable future it appears that the SECs recent dominance will continue.
With that said, I will say that, if you meet an SEC fan who tells you it's always been that way, you can know from multiple sources, that's just not the case. And if the Big 10 and Big 12, which were the dominant conferences not even a decade ago, start paying big money for the best coaches, they may catch up quite quickly.
Well, any thoughts? Hopefully, all of the tables are not too much data to digest all at once.