Playoff odds

Each remaining game is simulated by drawing both teams’ scores from a normal distribution fitted to their own weeks this season, then re-seeding on the league’s real rule (win %, ties broken by total points).

Each team’s average is pulled toward the league average before simulating. That is not a fudge: the raw model was checked against seven real seasons and was measurably overconfident — teams it gave a 30% chance made the playoffs 46% of the time. Shrinking the means cut the average error from 4.9 points to 0.8.

What it ignores: injuries, bye weeks, trades, waiver pickups, and hot streaks. Weeks are drawn independently. So treat a 95% as “very likely”, not a promise.