Wednesday, January 12, 2011

Ranking the Conferences

TSF is just a short ways away from publishing our final rankings for the 2010-2011 season. But before we get to that, I'm going to post a ranking of all the conferences that had some presence in the Top 25 season. For the purposes of this article, let's just call those conferences the Top-25'ers. Before I get into the messy business of ranking them, I have to explain that this makes for no exact science, and I need to give the background to the methodology I applied. In no way can we make an apples-to-apples comparison, but hey...it's fun to try.

First, as you know, every Division I school plays some lower tier opponents to soften their out of conference schedule. Some teams play more of these games than others. And some conference schedules allow for more intra-conference play that others. As a result the number of inter-conference play between Top-25'ers varies between one another. This means Top-25 conference X may play 33 contests against other Top-25'ers while conference Y may only play 20. I've taken that into consideration as best I can.

To compile the raw data, since I don't have the fancy database search tools at my hands as do the broadcasters' announcers, I had to manually assemble this looking through each team's results. Full disclaimer, I may have one to two things wrong due to human error, but I did the best I could not to make any mistakes. So onto the methodology...each conference has a row of data including wins, losses, total inter-Top-25'er contests, and win percentage. I have compiled six sets of standings. The first two have the same data as one another, just sorted differently. They contain the standings for the Top-25'ers sorted by win percentages and the same sorted by number of inter-Top-25'er conference contests. The second set of two holds the same data, respectively, but for the bowl season only. And finally the last two standings consider the entire season, regular and bowl.

First, let's start with the regular season.



We can see that the Big 12 took on the most inter-Top-25'er contests, and on top of that they finished with the highest win percentage. Trailing by just a little in win percentage, the SEC places right behind the Big 12, but with considerably less games. The rest of the Top-25'ers play roughly the same amount of games as the SEC, so we probably don't need to lean on the sorted by numbers of inter-Top-25'er contests standings too much here. After the SEC, we see a gap between the 2nd and 3rd place conferences. The Big Ten and Pac 10 take the next spots in that order and with similar win percentages, followed by the WAC and MWC. The ACC trails them, and the Big East is hanging in the basement.

While the number of inter-Top-25'er games doesn't necessarily tell us which conferences are stronger than others, it does tell us which programs are trying to put together exiting and challenging schedules, which of course factors into individual team rankings. The Big 12 is clearly making the biggest effort.

Now, let's turn to the bowl season.



Here, the win percentages can't sand on their own. The WAC is first for crying out loud, and even worse the Big East ranks higher than the SEC! So for the bowl season, let's weigh how many teams actually competed in bowls. I think this strongly represents people's perceptions of how these conferences compare to one another, with the exception of the Big East. They almost certainly don't deserve a standing higher than the Pac 10. But the Pac 10's low number of bowl bids shows how top heavy and lopsided the conference is. Although this chart doesn't reflect every bowl bid for every conference (some played against non-Top-25'ers), it does for the Pac 10. If we throw in USC pretending their scandal didn't exist, that wold still only bring them to tied with the Big East. Our ACC did okay in this respect. This chart reflects the inter-Top-25er bowl play for the ACC, but the total number for the ACC is actually higher with 9 appearances.

And finally, let's look at the entire season.



I weight the standing here with the sorting by win percentages more heavily. Even though the number of inter-Top-25'er contests vary between 33 and 20, 20 is still a sizable enough sample set to gage a win percentage that indicates something significant. In that chart the Big 12 just barely edges out the SEC. The Big 12 did so well in the regular season, but they had a very mediocre performance in the bowl season, but they still managed to hang on to their narrow lead over the SEC. The SEC pretty much held on to where they were in terms of win percentage, with only small drop. The Big 10 however, performed abysmally during the bowl season in comparison to the hype we heard all season long about the strength of that conference, but fortunately for them, only weaker conferences trailed them and had too much ground to make up from the regular season to catch them. All of those conferences behind the Big Ten made up some ground, but nothing dramatic.

So there you have it. That's how the Top-25'ers stack up to one another. We could use this for speculation into seasons to come, but that has its limits with some of the movement of programs between the conferences in coming seasons. TCU heads to the Big East. Utah and Colorado head to the Pac 10, err Pac 12. Nebraska heads to the Big Ten with twelve teams, leaving the Big 12 with, umm, ten teams. And who would dare say that doesn't make sense? And the SEC has accepted the James Madison Dukes...PSYCH! Comparing the conferences always provides a good barometer for how teams earn their rankings in the BCS, the poll, and of course right here on TSF. As I said earlier, it's not an exact science. Somebody else running this exercise with his/her own method could produce different results, although I'd get a little suspicious if the Big East or the WAC wound up beating out the SEC. But nonetheless, I think the endeavor worthwhile for a little speculative fun.

Something that I didn't measure was a consideration of the number of games that ended with the difference of a single touchdown or less. My raw data includes that, so maybe there'll be a follow-up on that. The argument on that has some given and take. While you could say that a game that close could have gone either way, we could also say that it's the elite programs that finish strong in the close games, hence making them elite programs. So does winning by seven or less mean a weaker win or a stronger team? I think there's a case for both.

You can access the raw data here: 2010_conference_rankings.ods. You will need Open Office to view it. If you don't have that program, it's freely available for all the commonly used platforms: Open Office. (If you run Mac OS, Neo Office will also work if you don't want to run X: NeoOffice)

Looking onward to the 2011-12 season...LET'S GO HOKIES!

No comments: