SCHEDULE DISPARITY · NET EDGE GAMES

Schedule Edge

Which teams a season's schedule favored, counted in games with a real rest edge and priced in wins. Not a prediction — much of the gap is structural, and each season stands alone.

HOW THIS IS CALCULATED

The market check — season win totals

If a schedule edge compounded over a season, the teams at the top of this page should beat their preseason win-total lines. Across 31 seasons of archived lines, they don’t:

Net edge gamesWent overpercent of the bucketTeam-seasons
+10 or better48.0%102
+5 to +946.5%144
−4 to +449.5%386
−9 to −545.4%163
−10 or worse47.7%86

No gradient, in either direction. The correlation between a team’s net edge games and its finish against the line is r = -0.02 across 884 team-seasons — statistically zero. The rest edge is real game to game, and that record lives on the Model Results page. But over a full season it amounts to a few possessions here and there, and the market’s win totals already price the schedule. This is a null result, published on purpose: season over/unders are not beatable from this page, and this site won’t pretend otherwise.

Lines from the sportsoddshistory.com archive (hosted at covers.com), 1993-94 through 2025-26. No lines were published for the 1998-99 lockout, and 2019-20 is skipped here because its season was suspended at 63 to 67 games — a preseason win total never got a full schedule to resolve against. 3 pushes are excluded from the rates; overs hit 47.9%overall — win totals lean under league-wide, edge or no edge. The archive’s win count matched this site’s own game records for every one of the 884 team-seasons before anything was computed.