AndersonSports BCS Rankings: the Anderson and Hester Football Computer Rankings

Why Is Missouri Ahead of Oklahoma?

By Jeff Anderson

(December 18, 2007)

We’ve gotten a lot of emails asking why #2 Missouri is ranked ahead of #8 Oklahoma. In a sense, this isn’t surprising. Conventional wisdom certainly suggests that a team that has beaten another, let alone twice, should be ranked ahead of the defeated team, provided that the victorious team has at least the same overall record. So why do we depart from this conventional wisdom? Why, as many of you have asked, don’t we put more weight on head-to-head games? (After all, we also have 2-loss Arizona State ranked ahead of 2-loss USC and 2-loss Virginia Tech ranked ahead of 2-loss LSU.) Well, here is our answer:

There is a difference between which team is better right now (or which is better head-to-head) and which team has had the better season-long results; between which team would be hardest to beat today and which has done the most on the year. The former is the team that the pundits or gamblers would pick to win right now; the latter is more deserving of season-long accolades. There are two different questions: “What are teams capable of?” and “What have they have earned?” And their answers represent two different things.

As our rankings page makes clear, we are only, solely, entirely, concerned with the second of these questions—with what teams have earned, on the field. We seek to reward actual accomplishments, not perceived potential. Our rankings are a list of which teams have the best records considering their schedules—nothing more, nothing less.

Some of you might respond, “How can you say Oklahoma hasn’t earned the higher ranking when the Sooners have beaten the Tigers twice on the field?” Our answer is simple: because the season isn’t only 2 games long.

Others have asked, “Why don’t you count head-to-head more heavily?” Our answer is as follows:

Every game is a head-to-head matchup. Oklahoma lost head-to-head to #46 (6-6) Colorado and head-to-head to #41 (8-4) Texas Tech. Yet we rank Oklahoma ahead of those teams, because in other games the Sooners have been so much better. In fact, we'd still rank the Sooners ahead even if they were to play one of those teams again and lose again. We doubt many college football fans would object to this.

It's the same thing with Missouri. Missouri has been so much better than Oklahoma in other games that the Tigers still rank ahead of the Sooners overall. How much better have the Tigers been in other games? Apart from playing Missouri, the Sooners have gone 9-2 with as many losses as games played (2) versus the top-40—hardly earth-shattering. Apart from playing Oklahoma, the Tigers have gone 11-0 with wins over #3 Kansas (the Jayhawks' only loss) and #13 Illinois—that is, a perfect 11-0 with a nation-leading 2 wins versus BCS bowl-bound teams.

With their 2 wins over Missouri, Oklahoma has very nearly caught the Tigers. Since they now both have 2 great wins, the central difference between them is that Missouri has lost twice to a top-10 team (Oklahoma), while Oklahoma has lost twice to non-top-40 teams. Because of this, Missouri clearly still has the better season-long results—just as Oklahoma would clearly still have the better season-long results if Colorado or Texas Tech were to beat the Sooners a second time. In fact, Oklahoma has been so much better than Colorado and Texas Tech in other games that—unlike with Oklahoma and Missouri—the Sooners could actually sustain 2 more losses to either of them (for a total of 3 losses to a single team) and still be ranked higher than the Buffalos or Red Raiders.

Again, every matchup is head-to-head, and you've got to look at all 13 (or 12, or whatever) games, and not just 2. If one team has beaten another—twice!—and yet still remains behind in season-long results, then this means that the other team has clearly been much, much better—yes, head-to-head—against everybody else. That is the case with Missouri and Oklahoma—as it would be the case with Oklahoma and Colorado or Texas Tech.

This isn’t any different than in other sports. If the Yankees were to finish at 100-62 and the Red Sox at 99-63, nobody would care whether the Red Sox had outplayed the Yankees head-to-head. Nobody would care even if the Red Sox had swept the Yankees every time the two teams had played. The Yankees would still have earned the divisional pennant. Rafael Nadal usually gets the better of Roger Federer when they play, but few doubt that Federer is the world’s #1 player. Why? Because head-to-head is not the only gauge of accomplishments.

True, Missouri’s overall won-lost record is not better than Oklahoma’s. But the Tigers’ won-lost record-considering-their-schedule is better. And this brings up another problem with artificially weighing head-to-head matchups. Missouri’s record, considering its schedule, is better than Kansas’s, which is better than Virginia Tech’s, which is better than LSU’s, which is better than Georgia’s, which is better than Arizona State’s, which is better than Oklahoma’s. If one were to flip Missouri and Oklahoma in the rankings, then, one could ask, how did Oklahoma somehow get ahead of 5 others teams that the Sooners would otherwise be behind? And how did the Tigers get behind 5 other teams that they would otherwise be ahead of?

As these questions indicate, there is no logical way to randomly assign additional value only to some head-to-head games, because every game is a head-to-head matchup. Nonetheless, general perception rightly points out that there is something important happening when two teams in close competition with one another play each other. For when making a direct comparison between two teams, head-to-head games effectively count twice as much as other games—and they do so naturally, as logical principles dictate. In baseball, the Dodgers can make up 1 game on the Giants in the standings in either of two ways: either by beating the Cardinals (or whoever) while the Giants lose to the Cubs, or by beating the Giants head-to-head. Similarly, in comparing Missouri and Oklahoma, the games that Missouri has played head-to-head against Oklahoma count exactly as much as any 2 games either team has played separately. In baseball and in college football alike (at least in our rankings), when comparing two teams head-to-head, their head-to-head games count exactly twice as much as other games—not more, not less. Missouri is still ahead of Oklahoma because, aside from Oklahoma’s 2 wins over Missouri, the Tigers were effectively more than 2 games ahead in the standings.

In conclusion, the only way a team can beat another team twice and still remain behind in our rankings is if the team that loses those 2 head-to-head games has been much, much better otherwise—and that is the case with Mizzou and OU. We are not saying we think that Missouri is better than Oklahoma right now (nor that Arizona State is better than USC, nor Virginia Tech than LSU). That is not what our rankings reflect. We are merely saying the Tigers have accomplished more on the season, on the field, to date. Despite losing twice to Oklahoma, they have still posted a better won-lost record considering their schedule than the Sooners have. Thus, based on their accomplishments, they have earned the higher ranking—regardless of who anyone thinks is the better team today.