What's the Best Conference in College Football?

1. Foreword
On Saturday, December 2 of last year, as people debated whether Florida or Michigan should be invited to face Ohio State in the BCS National Championship game, the perceived historical strength of the SEC and weakness of the Big Ten was taken as a given by some commentators and voters. This was not a new statement, and I'd heard it so many times before that I almost believed it. But as a lifelong fan of a Big Ten team, it irked me and I wanted to see for myself if it was true, or just another myth. To answer my own question, I've prepared a comparison of the BCS conferences over the past 15 seasons (and shorter intervals) and here I'll share with you my results.

2. Summary
While the SEC has probably been the toughest conference in college football over the past 15 seasons (relying heavily on its sterling bowl record over that period), in the past 5 seasons both the Big Ten and the ACC have been better. The ACC's 0.5416 winning percentage bests both the Big Ten's 0.5214 and the SEC's 0.5047 over that period.

3. Discussion

3.1 Methodology
Using the marvelous treasure trove of data available at jhowell.net and (since they weren't available there yet) the BCS conferences' own results for 2006, I stripped out the last 15 seasons for every team that currently plays in a BCS conference. I pared the data down further by then extracting teams' records against non-conference opponents from other BCS conferences (e.g. Florida State's record against the Big East, the Pac-10, etc.) and removing data against intra-conference teams and non-BCS foes/cannon fodder. I've also included Notre Dame in all of these records, since they've consistently been a good enough team to play in a BCS conference.

Teams are treated as if the conference they're currently in has always been their conference (so Louisville has always been a Big East team and Texas has always been in the Big XII). The purpose here is to find the strength of each conference based on the teams it presently consists of, not its historical roster.

The problem comes in with the Big East teams that jumped ship for the ACC. The way I've treated all other teams, Miami, Boston College, and Virginia Tech's conference games against other Big East opponents would now count as ACC victories over the Big East. However, the sheer number of games they'd bring with them against the Big East would seriously distort the results. Those three teams played 156 games against Big East opponents, while the rest of the ACC combined only played 263 games against all the other BCS conferences over that period.

I thought about just throwing out those games, but who's to say they wouldn't have at least played some portion of those games anyway? In the end, I assumed each of the teams would have played one game per season against a Big East team, and used that as a conversion factor. It's ugly, in that you end up with teams getting partial wins and losses, but I think it's fairer than giving them credit for everything or nothing.

I've also taken bowl games into account for this analysis. Even if some teams seem to rarely show up for their bowl appearances, you're at least supposed to play to win the games, and they can be a very significant portion of a conference's games against other BCS opponents.

3.2 Fifteen-Year Results
Fifteen seasons of non-conference data have been distilled into the following table.

**Table 1: 1992-2006**
	Total				Regular Season				Bowl
Conf.	W	L	T	Pct	W	L	T	Pct	W	L	Pct
Big East	90.8	180.2	3	0.3332	78.8	165.2	3	0.3210	12	15	0.4444
Big XII	134	146	1	0.4777	97	100	1	0.4911	37	46	0.4458
Pac-10	122	129	3	0.4832	93	100	3	0.4781	29	29	0.5000
Big Ten	195	173	3	0.5277	153	127	3	0.5435	42	46	0.4773
ACC	197.2	167.8	3	0.5381	155.2	131.8	3	0.5380	42	36	0.5385
SEC	148	117	1	0.5574	93	77	1	0.5455	55	40	0.5789

Over the past 15 seasons, the SEC has been ahead of everyone else. They've possessed both excellent regular-season and postseason records, although the regular-season mark is barely ahead of the Big Ten.

Once could also make the argument that the Pac-10 has been underrated through this period. There's a considerable drop-off between the top three conferences and the Pac-10/Big XII duo, but I would argue that the Big XII has gotten undeserved respect over that timeframe and the Pac-10 has been undervalued. They possess nearly identical regular-season records over that span, and the Pac-10 has been much better in bowl season.

Then, there's the Big East. It's really no surprise that their winning percentage has been this low. Over this stretch, teams playing in that conference not called Miami and Virginia Tech have gotten exactly the respect they deserve. The Big East teams now on the rise hadn't yet really started that rise back in the early to mid-nineties.

One more thing I'd like to point out when looking at this data is to take notice of the number of regular season games played out of conference. Distinctly lagging is the SEC, which supports another piece of conventional wisdom. The SEC really doesn't (or at least didn't) play anybody. The Big Ten played 283 regular-season non-conference games against BCS opponents. That's an average of 1.72 games per team per season. The SEC played 171, or 0.95 games per team per season. Whereas most Big Ten teams took on 2 BCS opponents a year, the SEC filled that slot with some scrubs from the Sun Belt. The Big XII also mostly stayed at home, though not to the SEC's extent.

3.3 Ten-Year Results
Looking at the 11-season and 10-season results, a turning point appears. In Table 2, the ACC has improved slightly and claimed the overall lead for toughest conference, but only by a slim margin over the SEC. The Big Ten itself isn't far behind at all. A swing of only two games in the SEC puts them ahead. In Table 3, that swing manifests itself. The SEC's 14-3 1996 season is no longer counted and the Big Ten drops its 19-14 season. Both conferences had above-average seasons, but the SEC's was more of an outlier. The ACC outperforms both and slightly increases its overall lead.

**Table 2: 1996-2006**
	Total			Regular Season			Bowl
Conf.	W	L	Pct	W	L	Pct	W	L	Pct
Big East	62.9	121.6	0.3409	53.9	108.6	0.3317	9	13	0.4091
Big XII	92	104	0.4694	63	69	0.4773	29	35	0.4531
Pac-10	94	94	0.5000	71	73	0.4931	23	21	0.5227
Big Ten	139	122	0.5326	105	86	0.5497	34	36	0.4857
SEC	110	94	0.5392	66	63	0.5116	44	31	0.5867
ACC	138.6	116.9	0.5425	109.6	91.9	0.5439	29	25	0.5370

**Table 3: 1997-2006**
	Total			Regular Season			Bowl
Conf.	W	L	Pct	W	L	Pct	W	L	Pct
Big East	54.3	100.3	0.3511	45.3	88.3	0.3390	9	12	0.4286
Big XII	84	96	0.4667	57	63	0.4750	27	33	0.4500
Pac-10	89	84	0.5145	67	65	0.5076	22	19	0.5366
SEC	100	91	0.5236	60	60	0.5000	40	31	0.5634
Big Ten	123	111	0.5256	92	78	0.5412	31	33	0.4844
ACC	121.3	100.3	0.5474	93.3	78.3	0.5437	28	22	0.5600

In other conferences, the Big East has made significant improvements and its winning percentage improves continually as we shrink the timeframe we look at. The Pac-10 ducks over the .500 barrier briefly, while the Big XII has slipped just slightly.

3.4 Five-Year Results
I consider the five-year data to probably be the most relevant to the argument of which conference is strongest today. The Big East isn't punished so badly for its programs that are just now making themselves into national contenders, and the established conferences don't get to coast on their legacies. At the same time, there's enough data so that small sample sizes aren't a big problem.

**Table 4: 2002-2006**
	Total			Regular Season			Bowl
Conf.	W	L	Pct	W	L	Pct	W	L	Pct
Big East	34.8	55.2	0.3863	29.8	48.2	0.3816	5	7	0.4167
Big XII	44	47	0.4835	30	30	0.5000	14	17	0.4516
Pac-10	45	47	0.4891	33	37	0.4714	12	10	0.5455
SEC	54	53	0.5047	35	38	0.4795	19	15	0.5588
Big Ten	61	56	0.5214	45	38	0.5422	16	18	0.4706
ACC	68.2	57.8	0.5416	52.2	46.8	0.5276	16	11	0.5926

The data show that he ACC has been the toughest conference over the past five years. I've checked, and even if you throw out all the former Big East vs. Big East matchups, that's still the case, though it does drop to a sub-.500 conference in the regular season. Their unchallenged record in bowl games pulls up their overall winning percentage enough to hold on to the top spot. That they maintained their lead despite how hideous the conference looked this past season shows how strong a conference it's been over the whole of the past five years.

The Big Ten is the clear winner here when it comes to regular season winning percentage, but the usual swoon in bowl season holds their numbers down. The surprise is that the Big XII performed so well in the regular season over this stretch. In the other time frames examined here, it's been a sub-.500 conference. Their bowl winning percentage still hasn't improved, however, so the Pac-10 remains the better overall conference.

One reason that the bowl record has such an impact is that the Big XII as a whole barely scheduled any BCS opponents. Each team played an average of 1.00 games against non-conference BCS opponents per season. Over the same stretch, Big Ten teams played 1.51 of those games per team per season, showing the decline of tougher scheduling in the north. On the other hand, SEC teams have improved to 1.21 games per team per season, bucking their historical trend.

3.5 Other Five-Year Spans
Breaking down the data in a different way, we can see that the conventional wisdom comes from what was once true: The SEC was the dominant conference, but they never played a decent non-conference opponent. If we examine the five-year results from 1992-1996 and 1997-2001, we should be able to illuminate some of the larger trends.

**Table 5: 1992-1996**
	Total				Regular Season				Bowl
Conf.	W	L	T	Pct	W	L	T	Pct	W	L	Pct
Big East	36.5	80.0	3	0.3097	33.5	77.0	3	0.2995	3	3	0.4444
Pac-10	33	45	3	0.4151	26	35	3	0.4160	7	10	0.4118
Big XII	50	50	1	0.4975	40	37	1	0.5161	10	13	0.4348
ACC	76.0	67.5	3	0.5238	62.0	53.5	3	0.5295	14	14	0.5385
Big Ten	72	62	3	0.5314	61	49	3	0.5471	11	13	0.4583
SEC	48	26	1	0.6443	33	17	1	0.6535	15	9	0.6250

It shows up clearly here. The SEC was dominant in the early and mid-'90's, but they never bothered to play anyone in the regular season. The Big Ten played more than double the number of equivalent games over that stretch, playing 2.05 per team per season, while the SEC only scheduled 0.85 per team per season. Alabama and Mississippi State each only played two games against non-conference BCS opposition. Arkansas, Auburn, and Ole Miss didn't play any. Any!

At the same time, they won a huge number of the games that they did play. A 0.6443 winning percentage across the conference is an impressive number that no one else can touch. They absolutely deserved to be known as the best conference in football.

Let's look at the next five years after that.

**Table 6: 1997-2002**
	Total			Regular Season			Bowl
Conf.	W	L	Pct	W	L	Pct	W	L	Pct
Big East	19.5	45.0	0.3022	15.5	40.0	0.2791	4	5	0.4444
Big XII	40	49	0.4494	27	33	0.4500	13	16	0.4483
Big Ten	62	55	0.5299	47	40	0.5402	15	15	0.5000
Pac-10	44	37	0.5432	34	28	0.5484	10	9	0.5263
SEC	46	38	0.5476	25	22	0.5319	21	16	0.5676
ACC	53.0	42.5	0.5551	41.0	31.5	0.5657	12	11	0.5217

Of course, my first impression is that the late '90's were not kind to the Big Ten, but if you look closer you can see that's not really the case. The bowl winning percentage crept all the way to .500, while the overall dropped by less than 0.15%. It's that the ACC and the Pac-10 came on like a house afire, while the SEC dropped back to Earth. I'm as shocked as anyone to see the Pac-10 riding so high, especially before USC's return to the college football elite, but there it is. Mid-level Pac-10 teams managed to win just a few more games than they lost, and there the numbers are.

**Table 7: 2002-2006 (Table 4, reprinted)**
	Total			Regular Season			Bowl
Conf.	W	L	Pct	W	L	Pct	W	L	Pct
Big East	34.8	55.2	0.3863	29.8	48.2	0.3816	5	7	0.4167
Big XII	44	47	0.4835	30	30	0.5000	14	17	0.4516
Pac-10	45	47	0.4891	33	37	0.4714	12	10	0.5455
SEC	54	53	0.5047	35	38	0.4795	19	15	0.5588
Big Ten	61	56	0.5214	45	38	0.5422	16	18	0.4706
ACC	68.2	57.8	0.5416	52.2	46.8	0.5276	16	11	0.5926

Taking a look once more at the most recent five years throws a few things into even starker relief. Number one on that list is that the Big East has gotten much, much better very quickly. To improve your conference-wide winning percentage by over 10% in only five years is tremendously impressive. They shouldn't be spoken of in the same breath as the power conferences just yet, but they're certainly not the doormats they used to be.

As USC has gotten better, the rest of the Pac-10 has actually gotten worse. They're still underrated, but the conference was actually healthier when its premiere member was still struggling.

The Big XII has gone the other way. They're still the fifth-place conference, and still overrated, but they're less fraudulent than they were in the late '90's. Led by the resurgence at Texas and Oklahoma, they've been a better overall conference.

4. Conclusions
At the beginning of 1997, the SEC was coming off of an impressive five year run, crushing, killing, and destroying all before them. Their impeccable regular season record was built around playing very few non-conference games against the future BCS conferences, and playing 60% of the ones they did schedule on their home fields. Ten years later, this situation is still accepted by some people as the truth, but things have changed. The ACC is the best all-around conference, the Big Ten is the best in the regular season, and their margins are both much smaller than the SEC's was. The Big East in the past five years has made up a lot of ground between "functional bye" and "contender", even if they aren't really there yet.

Finally, I'd like to again note the way scheduling has changed. The SEC's non-conference slate used to be a fraud, apart from a few rivalry games. They're still well behind some of the other conferences, but they play over 40% more BCS competition in non-conference play than they used to (1.21 vs. 0.85 g/p/s). The Big Ten has slid the other way. On average, half of its schools have dropped one of their non-conference games against a BCS school (down to 1.51 g/p/s from 2.05). The Big XII has taken over as the conference with the weakest non-conference slate. Each school plays about one non-conference game a year against competition that isn't from Directional State.

With the advent of the permanent 12-game regular season, I'd really hope to see this change. The Pac-10 plays nine intra-conference games, so they're exempt, but the other schools on this list should take it as an opportunity to play at least two of their non-conference games against quality opposition. Even if that means we see a lot more Ohio State-Baylor and USC-Indiana games, it's better than the sham that is filling your out-of-conference schedule with Eastern Michigan, Florida International, and Nicholls State. Yeah, I'm looking at you, Wisconsin. And you too, Clemson and Alabama. Don't think I don't see you over there, hiding behind Louisiana-Monroe and Temple. Wusses.

<-- Back to the HSR