Sure, this is a good, safe consensus line of thinking.
But even for conference tie-breakers, at least for the top of the conference, you're putting h2h above worst loss. If Team A and Team B are tied for the top of the conference and Team A beat Team B h2h, then by definition, Team A also has a worse loss. Team B lost to a conference co-champion. Team A lost to some other, lesser team.
That's what I mean by "mattering" more than the "mere 1-game amount" of 8.3%.
If Team A's loss is to the 3rd-best team in the conference, the general consensus thought would be that it's not an issue and default to h2h outcome.
But how bad does the loss have to be to "matter" more? If Team A lost to a .500 team, is that bad enough to not "overvalue" h2h? If they lost to the worst team in the conference? Where's the line?!? Is there one?
I know outcomes should matter, but shouldn't all outcomes matter equally? Isn't one outcome out of 12 just simply one outcome out of 12? Isn't each outcome simply a snapshot of an individual game on a given Saturday (or any of the other days lol)? Hence, all of us agreeing that the transitive who-beat-who-beat-who yielding some sort of terrible Kent State deserving the national championship some years as absurd.
I just like to have this conversation logically explored.
This is where things get really dicey, though, as I said. Let's assume three teams:
- Oregon finishes 12-0, 9-0, does not play Ohio State or Indiana.
- Ohio State finishes 10-2, 8-1, with narrow losses to Texas (finishes 11-1) on the road in week 1 and Indiana at home in week 11. They steamroll the rest of their schedule and beat Michigan by 3 TDs in the final regular season game.
- Indiana finishes 10-2, 8-1, with a narrow loss to a 7-5 MAC team in an uninspired performance in week 1, generally competent but not dominant wins the rest of the way, then beats OSU on the road in a game where they are +2 in TO margin. Then they go on the road to Purdue and get shellshocked in a game that everyone thought they were going to win.
- We'll assume no other B1G team finishes 8-1 in conference play so OSU/IU is the tiebreaker for the CCG.
In the rankings, and in the CFP seeding, I'd have no problem with Ohio State being above Indiana.
But for CCG inclusion, what *OBJECTIVE* and mathematical way do you have to justify Ohio State getting the CCG slot over IU? If we're talking eye test, if we're talking "quality wins" vs "bad losses", I totally get it.
I just don't know how I can consider those things to trump H2H when it comes to CCG selection. We have
objectively defined tiebreakers for a reason. The reason is that if we're trying to make it an "eye test", who decides? Conference athletic directors that are financially incentivized to want an Oregon/OSU CCG for ratings purposes? Or conference athletic directors that are financially incentivized to want as many teams as possible in the CFP and who know that OSU with narrow losses to quality Texas, Indiana, and Oregon teams may still get into the CFP, while IU will still likely get in by being 10-2 but likely not 10-3 with a stomping by Oregon? Do you engage the CFP committee to make the decision?
And then when you put OSU in, how do you blunt the criticism of all the people who look at that H2H result and say "scoreboard!"? Even if you find a way to put OSU in, you'll piss a LOT of people off and they'll wonder if the games even matter...
I see where you're coming from. I just don't see a particularly tractable solution.
