When the College Football Playoff was first conceived in 2013, the Playoff and the major conferences decided that a four-team format was best suited for including the "four most deserving" teams in the country, regardless of conference.
Since there are five power conferences, there couldn't be an auto-bid system, and this theoretically left the selection committee some room for discretion if it felt that two teams in the same conference were far and away two of the best teams in the country, as was the case in 2011, when the Bowl Championship Series computer system named two SEC teams, LSU and Alabama, to the national championship.
In the two years of the Playoff's existence, the committee has yet to face this kind of decision, because the choices have been fairly clear-cut and spread across the conferences. This season could be the first real test of the committee's willingness to reward the four best teams, and not just the best of the conference champions. After three weeks, two ACC teams (No. 3 Louisville and No. 5 Clemson) and two Big Ten teams (No. 2 Ohio State and No. 4 Michigan) make up four of the top five spots in this week's AP poll, and it stands to reason that even with a loss, each of those teams could have a very good claim to a Playoff spot.
First, take the ACC, the conference with perhaps the best chance of landing two teams in the Playoff. Coming off a 63-20 rout of Florida State, the Cardinals are destroying everyone in their path. And with a tough but manageable schedule, Louisville could plausibly gain the top spot in the Playoff by winning out. That résumé would include wins over Florida State, Clemson, Houston, and the ACC Coastal champion (say, North Carolina).
But what if Louisville has a loss to undefeated Clemson? If the rest of the major conferences have undefeated champions, then the Cardinals will likely be out of the Playoff, but the Big 12 has virtually no chance of having an undefeated champion. The Pac-12, with Stanford, Washington, Oregon, and UCLA, has a better chance of beating itself up than having a clear-cut top team.
So in that case, why not Louisville?
Now, let's say Washington wins the Pac-12 with wins over Oregon and UCLA but a loss to Stanford. Is that résumé better than Louisville dominating its schedule, with the exception of a loss at Clemson, and finishing with wins over Houston and Florida State? Probably not. If the Pac-12 champion has two losses, then it certainly shouldn't get in over a one-loss Louisville.
The same is true in the Big Ten. What if Michigan beats Ohio State to finish undefeated to OSU's 11-1? A Buckeye résumé that would include a dominant win over Oklahoma and a win at Michigan State could be better than that of another one-loss conference champion.
The question may come down to just how much weight the committee does put on conference championships, and how committed it is to choosing teams from different conferences, even if that requirement isn't written down.
In a season as wild as this one, a one-loss, non-conference champion might not even be a controversial selection. But will the committee be willing to pick a one-loss team with a superior résumé over a one-loss conference champion if it means a lack of conference diversity in the Playoff? We just might find out.