To be clear, I’m not saying there were *not* differences to defect across demos (clearly there were) but some of these demo comparisons can appear at least somewhat exaggerated because of this.

Some of this is explained by mathematical ceiling effects. If you were winning 90% of Group A last cycle and 30% of Group B, and then 10% of your existing supporters in both groups defect, you’ve fallen 9 pp in A and 3 pp in B even though there was no difference in propensity to defect across group.

I tried to tell all my anti-Trump friends skeptical of his chances in 2024, “man, if yall could see/hear the taxi and uber drivers here in NYC since 2021-2022 you would not be so skeptical”

8 each so far

I’m talking about using your power ratings (or FPI or other similar) and assigning W/L probabilities to each opponent for a typical CFP team on each team’s schedule, and then assessing difficulty for achieving a certain W/L record against that schedule. “Strength Of Record.” SMU 11-2 comes out ~15th

I was initially referring to pure strength of record (difficulty for a given CFP caliber team to achieve a given record against a given schedule). Using ratings such as yours indicate SC 9-3, Bama 9-3, BYU 10-2 all > SMU 11-2. That is fully counting ACCG though.

Is that a prediction of the committee’s behavior, or a pure analytical form of strength of record?

Understand the argument to simply pretend their Clemson loss never happened as a maneuver to protect the perceived value of qualifying for CCGs, and that doing so makes their resume more competitive analytically. But think having to use that (debatable) argument means it falls short of “OF COURSE”.

Completely ignoring today’s game, don’t your own ratings (and every other similar system) show that SMU did not have one of the 12 best/most deserving resumes coming into today?

Clemson (despite not having a CFP-worthy resume) and SMU (despite having a resume analytically ~equivalent to 3-loss SEC bubble teams) both got a chance last night to play their way straight into CFP quarterfinals on a neutral field. Hard to argue either is getting “punished” by “having” to play.

They’re really selling out to stop our run game. We’re gonna have to sit back and make some throws downfield. Guys should be open (and have been open).

A significant amount can be explained by demographics, perhaps even most of it. I suspect a disproportionate national + local media / social media focus on NYC-specific crime and disorder contributes here too but hard to quantify. Cost of living probably plays a role as well, as has out-migration.

My proposal is make every round a 1-2 game series, where the higher seed hosts the 1st game and lower seed 2nd game. No ET or PKs in either. The higher seed secures advancement by simply matching the 1st game result in the 2nd game, otherwise lower seed advances. No 2nd game if high seed wins 1st.