In the 10 years of the four-team College Football Playoff era, we learned two things rather definitively:
1. It's hard to rank teams.
2. A committee doesn't really do it any better than the BCS formula did.
The College Football Playoff committee has had a bit more margin for error than the BCS ever did, as it could select two more teams to join the party. In a decent percentage of this 10-year CFP experiment, this margin for error came in handy.
The selections were pretty easy in 2015, 2017, 2018, 2019, 2021 and 2022. In 2016 and 2020, things were a little blurrier, but the choices the committee made were fine. In 2016, it had to choose between two-loss Big Ten champion Penn State, which had lost to a decent Pitt team and gotten its doors blown off by Michigan, or one-loss Ohio State, which had only lost to PSU. The committee went with the latter. In 2020, it went with a one-loss Notre Dame team that had split two games with CFP participant Clemson instead of a one-loss Texas A&M that had gotten thumped by top-seeded Alabama. (It gave no consideration that year to an unbeaten and clearly awesome Cincinnati, but that's another story we'll get to below.)
That's eight of 10 years with easy or only reasonably tricky decisions. Even if the weekly rankings process sometimes exposed some inconsistency and contradictions in logic, again: It's hard to rank teams. You're always going to contradict yourself at some point. The committee gets a pass for that.
In 2014 and 2023, however, the committee was presented with genuinely complicated tests. It got a C-minus on the first test and an F on the second.