clock menu more-arrow no yes

Filed under:

Balancing “Who” vs. “How” is the committee’s greatest challenge

New, 55 comments

This is the crux of the debate over how to rank college football teams.

NCAA Football: Southern Methodist at Memphis Justin Ford-USA TODAY Sports

“They ain’t played nobody.”

The funny quip that’s often interjected sarcastically into college football discussion has old roots in the sport. A fan of a one-loss team from the SEC is sure to believe that the undefeated team from some other league doesn’t pass the smell, er, eyeball test, while the fans of the undefeated team shout “scoreboard!”

Fortunately for all fans of the sport, some very intelligent people have created formulas that adjust for strength of schedule, remove garbage time and analyze things like success rate and explosiveness on both sides of the ball. These offer a great idea of who has consistently performed at a high level in the games, that goes deeper than simply wins and losses. Matt Hinton, stats guru who formerly worked for CBS, has a nice summary of this season’s results below.

Being that this is an Alabama site, and Alabama has been at the center of more of these debates than any other team, we will focus on Alabama this year as a case study.

As you can see above, the consensus of the advanced metrics, or “baseline,” is that Alabama is a top 4 team this season. The Tide have been more consistent in the things they can control, play on the field relative to the other teams that their opponents faced, than all but three teams in the nation.

Yet, the playoff committee has them ranked 12th while Hinton has settled on a more defensible 9th. This is based on a résumé ranking that really looks at only one thing: who did you beat?

This is where college football discussions often start to resemble modern political discourse where two sides argue past one another, neither fully considering the other’s position. When looking at the Tide, they have beaten only two Power Five teams with winning records this year: Texas A&M and Tennessee, both 7-5. On the surface, that isn’t much for the Tide to hang their hats on.

Now, let’s compare Alabama’s résumé to that of Utah, currently ranked #5 and the team that will likely have the first chance at a playoff spot should LSU defeat Georgia. The Utes have three wins over Power Five teams with winning records, plus a win over 7-5 BYU, who defeated Tennessee early in the year. By this lone metric, they have accomplished significantly more than Alabama.

Of course, when comparing losses, Utah fell to a USC squad that is currently ranked #22 after playing much better down the stretch than earlier in the season, when the win over Utah was sandwiched between losses to BYU and Washington. Alabama has lost twice, to #2 LSU and #11 Auburn. All three of the losses were close.

Going back to Hinton’s compilation, the Utes check in at #7 on the baseline, with a pretty significant gulf in the ratings compared to Alabama and #5 Georgia.

So, how does this fit together? It really comes down to what’s more important: who a team has played vs. how a team has played. The latter is far more controllable than the former. Better stated: if Alabama played Utah’s schedule, and Utah played Alabama’s, how might the results change? Shouldn’t any contender beat up on average-ish Power Five teams, which says here includes the 8-4 Trojans?

Also, which is better evidence of a team that is good enough to compete with the top teams in the country: a close loss to unbeaten LSU, or four wins over those middling teams?

This is where advanced, opponent adjusted metrics come in. Bill Connelly is always quick to point out that his SP+ formula is intended to be predictive, but that’s kind of the point and it is based on the only data available, which is how the team has performed in its games. Back in 2015, Michigan State upset Ohio State in the Big Ten championship game, and was thus rewarded with a playoff berth despite finishing only 20th in SP+. Unsurprisingly, they were destroyed 38-0 in the semifinals by Alabama while Ohio State finished the year #1 in the SP+ metric and would have likely been much more competitive. Thanks for making Urban Meyer drown his sorrows in pizza, Sparty.

During the regular season, teams have less time to prepare for one another and things like short term injuries have a greater impact on the games. Would Alabama have defeated LSU if Tua Tagovailoa had been able to scramble, as he did the previous season when he ran for a 40 yard TD against the Tigers on an injured knee that still afforded a bit of mobility? Better put, might that game have been different if Joe Burrow had been the one hobbled by an ankle and unable to run?

To be clear, I’m not making a case for Alabama to be chosen for the playoff. Flukes and mitigating circumstances aside, the object of the game is in fact to score more points than the other team, and Alabama has failed to do that twice. When only four teams are allowed in out of 128, that will usually be a disqualifier, and I’m fine with that. The playoff is intended to crown one team, and no more than four are needed to get there. In fact, we only need a fourth team this year so that we can complete a playoff field as there appears to be only three worthy candidates.

Still, going forward, the committee should first consider how a team has played, and go well beyond just game scores. To date it appears that they are refusing to consider anything beyond the most rudimentary statistical data, and that’s disappointing. The good stuff is readily available to them and the devil is in the details. In this case the devil is boring, lopsided semifinals that are bad for the sport.