ESPN’s latest FPI Rankings leave much to be desired

Is the goal still to score more points than the other team?

Earlier this year I took some time looking at ESPN’s FPI rankings and had a lot of questions.  Although the latest FPI rankings now at least have Georgia ahead of Clemson (unlike right after they straight-up beat them), I’ve still got questions.

Georgia is atop the rankings while Alabama, Ohio State, Oklahoma, Michigan, and Cincinnati come in 2-6.  You can make some argument for an unbeaten Oklahoma over Ohio State, who lost at home to Oregon, but it’s hard to be overly upset about any of those teams.

So you’re aware, Notre Dame checks in 18th which is hilarious when you see some of the teams ranked ahead of them.  I obviously think they should be higher than that but as you’ll quickly see, losing closely seems to reward teams more than winning narrowly does in this formula.

So which teams had me the most bumfuzzled (shoutout to Bobby Bowden for that one) this week?

Here are just a few observations:

If Clemson beats LSU, the Colley Matrix will have Ohio State at No. 1

If Clemson beats LSU in Monday’s National Championship Game, then the Colley Matrix computer ranking will have Ohio State at No. 1.

In many ways, the Colley Matrix is my favorite of the major computer rankings out there. It’s not the results per se that make it wonderful, but it is–and I’ve been saying this for years–the best pure wins-and-losses strength of schedule rankings out there. I don’t use Colley for where it ranks the teams; but I will always cite its SOS numbers, because they’re the best.

There are two main reasons why I appreciate Colley more than any other ranking. The first is that his formula is public. On his website, Colley explains the rankings in both the pure math terms, as well as the logic behind them. That way, even if you’re not fully up on the precise numbers, anyone can really understand the basic concept behind how he gets his numbers.

More importantly, Colley only cares about wins and losses. Home, road, margin of victory, etc.–none of those things matter. It’s who you’ve played, and who they’ve played. Additionally, his matrix idea behind how to properly judge just how valuable an opponent is worth is as simple as it is innovative.

So, to repeat, Colley is the best system out there, in my opinion, for strength of schedule. As for the actual rankings–since the system ignores things like margin of victory and how well teams play, the results can get a little wonky sometimes. That’s okay; it’s still the best SOS system out there.

Which brings us to this year. According to Colley, the Ohio State Buckeyes have played the toughest schedule in the country this year. The only other teams with SOS numbers particularly close to Ohio State’s are South Carolina, Wisconsin, and Auburn. LSU has a pretty strong SOS. The Bayou Bengals sit at No. 14, and that will jump into the Top 10 after playing Clemson. Clemson’s SOS, however, sits outside the Top 50–and that’s after playing Ohio State. The number will improve further after playing LSU, but will still not be anywhere close to Ohio State’s.

Just how much harder is Ohio State’s schedule than Clemson’s? Well, the final rankings will have Ohio State ahead of Clemson, even if Clemson beats LSU. that doesn’t mean that Ohio State is better than Clemson this year or deserves a national championship; it does, however, speak to the fact that Ohio State’s schedule was significantly tougher than Clemson’s. Colley’s system prioritizes wins and losses over everything. The fact that Ohio State can stay ahead of Clemson with two fewer wins and one more loss tells you just how much higher Ohio State’s SOS is.

Some people may remember that Colley had UCF at No. 1 at the end of 2017. In fact, because Colley only cares about wins and losses–the system doesn’t value the National Championship Game more than any other game–it is the major computer that most often puts a team that didn’t win the NCG at No. 1. For example, its formula also had Clemson at No. 1 in 2015 and Notre Dame at No. 1 in 2012.

I’m not saying that the Buckeyes should declare themselves national champions this year, even though the NCAA would recognize such a claim (if Clemson wins the NCG). I am saying that it’s noteworthy how much harder Ohio State’s schedule was than Clemson’s this year, and these numbers back it up.

(As a complete tangent to end this post, it is a bit of a myth that Colley is the reason that UCF claimed the 2017 National Championship. UCF Athletic Director Danny White declared his team National Champions before that season’s National Championship Game was played. Had Georgia won the NCG that year, the Colley Matrix’s final rankings would have had Georgia at No. 1 and UCF at No. 2. It was only Alabama’s major second-half comeback that put UCF at No. 1 in the Colley Matrix, but the school didn’t care; it was declaring itself 2017 National Champions regardless. Had Georgia won, no NCAA-recognized selectors would have had UCF at No. 1; since Alabama won, there was exactly one–Colley–that agreed with UCF’s claim. But UCF was making that claim no matter what.)