The Conference Fallacy

If you're like me, and roughly half of your Facebook friends went to Notre Dame, then you've probably seen this article from Pat Forde.  If you haven't seen it, then take a moment to read it over.  In the article, Forde uses a reasonably objective methodology to determine which team had the most impressive resume in college football this season.  He comes to the conclusion that Notre Dame's schedule is stronger from top to bottom than that of Alabama or Georgia, and thus Notre Dame has every chance to win the BCS title game.  For those of us that have followed college football and Notre Dame closely, this makes sense.  Notre Dame has been far from perfect, but it would take a pretty good leap of logic to say they're far inferior to the other top teams.

Of course, there are others that are willing to make that leap.  Forde's article was in response to the assertion from a writer on an SEC blog that Notre Dame would be the seventh best team in the SEC.  That statement seems to come from the bottom of this post.  I don't see an objective evaluation* anywhere in there, so I'm not exactly sure what logic was used.  The only thing that his rankings say are that they come from "what (they) saw on the field."

*Just for giggles, here is an objective evaluation of the relative strength of the top six SEC teams versus Notre Dame.  I know Forde runs through the schedule in his article, but it's not exactly the most scientific evaluation in the world.  For my evalution, I will use three publicly-shared rating systems that use three different inputs.  They are:

Jeff Sagarin's predictor - This is based on points scored and allowed
Brian Fremeau's FEI - This is based on drive data
Bill Connelly's S&P - This is based on play by play data

Using these three different methodologies, we get the following rankings for the teams in question (note: we don't have FEI and S&P data for the last week yet, so these may change a little):

As you can see, the only team that is clearly ahead of the Irish is the Crimson Tide (Sagarin would favor Alabama by 6 on a neutral field).  You could probably make an argument for Florida and Texas A&M if you really wanted to as well, but the other three teams are clearly below ND.  This isn't to say that ND would "definitely" beat those three teams, but if they played them 10 times on a neutral field, I'm betting that ND would win 6 or 7 of them.  If Mr. Travis is looking for a team that would actually be the seventh best in the SEC, perhaps he should go with Utah State, who currently has an average rating of 19.33.

Understandably, a lot of people take issue with Mr. Travis' conclusion.  I do as well, but it's not that big of a deal to me.  It is an SEC blog after all, so it only makes sense that they're going to be a little biased towards the SEC.  Furthermore, if you find yourself reading that site regularly, then you probably wouldn't be expecting the world's most balanced and objective reasoning.  I imagine instead that the blog's readership is looking for a humorous take on SEC football.  While I don't have an issue with the blog in general, it does point to a problem that many college football analysts* have.  That problem is what I would call "the conference fallacy."

*I use that term loosely.

The conference fallacy is quite simple: People are too quick to judge a book (team) by its cover (conference).  Yes, you can make some accurate generalizations about conferences, but if you want to truly analyze which teams are the best, then you need to dig much deeper than that.  The Big 10 and SEC have long been known as the most traditional oriented conferences, but Urban Meyer has had great success with his version of the spread, and others (most notably Joe Tiller at Purdue) have opened up pass-wacky attacks.  The Big 12 is thought to be an offensive crazy show, but Texas last year and Kansas State this year have shown how a top notch defense can keep a team competitive and even superior to the rest.  Non-BCS teams are thought to be "finesse" teams, but BCS victories by TCU, Boise and Utah show how competitive the best of this group can be.  As conference membership grows ever more arbitrary, the differences between teams within a conference appears to grow and grow.

To add to this problem, the advent of super-conferences means that teams within a conference can have very different experiences.  For example, look at the differences between Florida and Georgia this season.  While they both played South Carolina and the rest of the East, they had very different experiences with the West.  Florida's cross-division opponents were LSU and A&M.  Georgia, on the other hand, played Ole Miss and Auburn.  Partially as a result of these differences, Florida's strength of schedule ends up 13th, while Georgia's is 42nd.  Florida's stronger schedule doesn't necessarily mean they were the better team, but it does show how not all "strong SEC schedules" are created like.

The moral of the story is simple:  If you're arguing about college football with someone, dare yourself to avoid mentioning conference affiliation when discussing teams.  This way, you'll be able to avoid the pitfalls of the conference fallacy.  And then eventually you'll realize that Notre Dame is pretty good at football.