Monday, November 10, 2014

The Biggest Problem With College Football

Bleacher Report recently posted an article entitled "How Would Florida State Fare If 2014 'Noles Played in SEC?". Yeah, this is the kinda crap that makes it clear you need a full playoff in the FBS. This article presents us with pretty much all that is wrong in how college football currently determines its national champions. First they focus pretty much entirely on what the computer rankings say about defense, ignoring the fact that Florida Sate has a better offense than all but 1 SEC team (And possibly better than that given that Winston sat one game). Further, they forget that those same stats show the ACC is nothing more than a small step behind the SEC in conference strength. In addition, Florida State plays a MUCH tougher non-conference schedule than almost any SEC team (How many play two Power 5 teams, let alone ONE one in their 4 non-conference games). 

They say that since 4 teams are ranked higher in the computer rankings Florida State would have already had 2 losses (and we know the higher ranked teams always win, like when #47 Texas A&M beat #5 Auburn, or when #41 Florida blew out #13 Georgia). And they commit the worst of all sins in sports (the transitive fallacy) when they compare Georgia's game against Clemson to Florida Sate's game (without Winston) against a much improved Clemson. Hey, didn't the team that just took down Auburn almost lose to ULM last week (So Auburn and ULM must be of a similar skill set)? Further (and this cannot be stated NEARLY enough), these computer rankings have a significant amount of uncertainty. CFB teams play far too few non-conference games for these computers to get nearly enough data to compare teams from different conferences. And most of those games come at the beginning of the season, meaning any data from those games is practically worthless by the end of the season. College level athletes are far from consistent. Coaches make changes. Players get injured. Young athletes improve their game. Some teams get lucky breaks (Auburn v. KState and LSU v. Wisconsin). Any ranking system that fails to account for this should be taken with a huge grain of salt (which is, realistically, all of them). Humans cannot factor all this in. And if computers cannot rely on this to compare teams from different conferences, what else what can they rely on? This is why we absolutely need a larger playoff with automatic bids by Power 5 conferences. And it is also why this "what if Team A played in Conference B" bullshit should be dismissed as the uninformed speculative garbage that it is.