In the aftermath of the New Years’ Day bowl games the airwaves were rife with unsubstantiated commentary from the “experts” who get paid to tell us what we should think we saw on TV. There’s not enough space on the Internet to correct all the misinformation that flowed our way yesterday, so I’ll pick just one particularly interesting topic: Where should UCF have been ranked before their exciting victory over Auburn? Should the Knights have been in the playoff? And, where should UCF be ranked now?
Yesterday we heard from people who think that by defeating the team (Auburn) that defeated two former number one teams (Alabama and Georgia) UCF proved itself best in the country. We heard from others who quite simply want to hand that title to the only FBS subdivision team to finish the season unbeaten. These people are a throwback to the bad old 1990s. If you have read this blog during the season, you already know why won/lost records and head-to-head results have no value when determining how good a team is.
The comment that deserves our attention came from a person in a position to exert his influence on the game and change the course of history: CFP Executive Director Bill Hancock. In the face of social media reaction to UCF’s victory and its ranking by the CFP selection committee, Hancock explained that UCF had been properly ranked at #12 because of its Strength of Schedule (SOS), implying that UCF played a weak schedule. This explanation is confusing since the CFP organization admits on its Website that it has “no specific metric” to measure strength of schedule. So how did the members compare the schedules? What information is Mr. Hancock referring to? Is it an assumption that a Group of Five schedule must be weaker than a Power Five schedule?
The Relative Performance Grading system (RPG) does have specific metrics for determining strength of schedule. The RPG metric includes not only a measure of opponent strength but also an adjustment factor for home field advantage (away game disadvantage). Using the RPG metrics, UCF did indeed play an easier schedule than all other top twenty teams. Notre Dame played the hardest schedule this season and that schedule was precisely 6.8% more difficult than UCF’s. You might be interested to know that Alabama’s schedule was precisely 3.39% harder than UCF’s.
Notre Dame and Alabama may have played a team or two that were better than anyone UCF played, but when comparing the total schedules, there’s not as much difference as one might think. It’s not like UCF played only cupcakes and the Power Five schools played only top fifty competition. Every Power Five team played its share of non-conference cupcakes as well as many soft conference members, e.g. Kansas, Baylor, Oregon State, Rutgers, Tennessee, Vanderbilt, Arkansas, North Carolina, Nebraska, and Colorado. Assumptions about conference strength simply mislead fans.
Another way to view the CFP committee’s ranking of UCF is to compare UCF to Wisconsin, a team that committee members admit played a weak schedule. According to RPG, Wisconsin’s schedule was precisely 2.7% more difficult than UCF’s schedule. If UCF’s schedule didn’t justify a higher ranking, then why did Wisconsin’s? Before the conference championship games, the committee ranked unbeaten Wisconsin at #4 and unbeaten UCF at #14. After UCF won its title and Wisconsin lost, the Badgers fell to #6 and UCF rose to #12. That feels like the same mechanical movements we used to get from the opinion polls in the 1990s: demote losers and backfill with winners; no thought required.
By comparison, RPG ranked UCF at #5 and Wisconsin at #9 prior to their conference championship games. RPG ranked UCF higher because it had played better football against nearly equivalent competition. Following the conference title games, Wisconsin dropped to #10 and UCF dropped into a tie for #6. UCF won and yet fell in the RPG rankings because Georgia and Clemson leapfrogged them with higher graded performances against better competition. You might remember that UCF gave up 55 points to Memphis and got a bit lucky to win in double OT. In other words, real thought went into the RPG rankings. UCF’s win was not automatically considered equivalent to those of Clemson and Georgia.
Going into its bowl game, UCF was ranked #6(tied) by RPG and its opponent, Auburn, was ranked #9. Auburn was ranked #7 by the CFP committee so many fans were influenced by the committee rankings to consider UCF’s win to be an upset. The RPG system did not consider it an upset, but rather a close win by a higher ranked team.
Now that all the games have been played, UCF has risen to the exalted #4 position in the RPG rankings but that doesn’t mean they should have been in the playoff field. Their new ranking is partially a result of losses by Clemson and Washington. Had UCF’s early season game against Georgia Tech not been cancelled due to a hurricane, the Knights may have added another power Five win to its resume and may have reached the top four in the RPG rankings. But, it is doubtful that a win over Georgia Tech would have propelled them from #12 to #4 in the CFP committee rankings.
To see all of our blog posts and all of our rankings, go to http://nemosnumbers.com/football-rankings/.