I am working on a player-based projection system, similar to how Bart Torvik does it. For now, the main projections are team-based, with the main inputs being:

1. Reversion to conference mean.
2. Coaching change factor = a fired coach results in a positive benefit, a good coach hired away produces a negative one.
3. Generic increase in performance for remaining players on roster.

It's okay, and generally mirrors some of the early subjective predictions. But we can do better.

Louisiana

No roster available for Louisiana