Notoriously variable lineups and literally a different starting pitcher in every game of every five game cycle) isn't your best work. It'd be kind of wild if football teams rotated 4-5 quarterbacks.

I do get your point that it's drawing on a wider, longer body of data and weighting a single outcome too heavily is disproportionate -

My point is that it's also disproportionate and distorting to weight previous seasons too heavily, esp in the NIL era when teams are getting blown up on an almost annual basis.

If the Padres beat the Dodgers on April 15 - sure, whatever. They've got another 15 matchups on the books. Every 18 or 12 game baseball pairing typically results in wins and losses. I mean, the White Sox destroyed the Yanks like 10-2 or something just last month. Baseball is a different animal.

In CFB, with very rare exceptions, you get one bite of the apple.