I’m traveling home today and had some time. I added a lot of data into ChatGPT including: average EPA (offense and defense), statistical averages, efficiency %s, strength of schedule, strength of victories, remaining schedule, etc.
To summarize: On average teams following a bye week have a 5-10% increase for EPA per play. Teams with a post- Week 12 bye have a 10-15% increase of EPA per play on average. Teams with a late season bye week on average have a 15% greater chance of victory against opponents than they would have prior to their bye week.
I asked to simulate the remaining games and this was the results:
Carolina 10-7
Bye
@ New Orleans (W 27-14)
Vs Tampa Bay (W 23-20)
Vs Seattle (L 31-17)
@ Tampa Bay (W 17-14)
Tampa 9-8
Vs New Orleans (W 27-17)
Vs Atlanta (L 27-21)
@ Carolina (L 23-20)
@ Miami (W 21-14)
Vs Carolina (L 17-14)
It references Carolina’s higher EPA per play following his the bye week, Tampa’s 3 consecutive divisional games, and apparent linear regression of EPA per play since season start as the reasons.