Strength of schedule is a measure of how difficult a team's opponents are. The NFL's official version, used for draft-order tiebreakers, is the combined winning percentage of a team's opponents from the previous season. The analytics version uses opponents' power ratings or point differential instead, because last year's records are a weak guide to this year's teams. The two can disagree by a dozen spots in the rankings.
Updated Sept. 16, 2026.
The league schedules every team against its own division twice, one full division from each conference, and three games against teams that finished in the same divisional slot. That last rule is why the official strength of schedule, computed from last season's records, punishes good teams: a division winner draws the other division winners. Offseason articles ranking schedules from easiest to hardest are almost always using this number.
The number has a problem: last year's records do not predict this year's. Roughly half the playoff field turns over each season. A schedule that looks brutal in May because it includes three teams that went 12-5 can be ordinary by October. Using preseason power ratings, or the betting market's implied win totals, gives a schedule difficulty that has some predictive value; using last year's win-loss gives one that mostly does not.
In-season, strength of schedule matters for adjusting the numbers: a team's raw EPA per play against three bad defenses is not the same as against three good ones, and opponent-adjusted efficiency is the fix. It also matters for the playoff picture, where the remaining schedule sets the projected win total.
Entering 2023 the Chiefs had the league's toughest official schedule at .578 opponents' winning percentage, built on the Bengals, Bills and Eagles' 2022 records. By rating-based measures it was middle of the pack, and Kansas City won the Super Bowl. The official number was measuring 2022.
Official SOS = the combined record of a team's 17 opponents (wins / games), from the prior season. Rating-based SOS = the average power rating (or point differential per game, or market win total) of the 17 opponents, weighted equally, often reported as the average expected margin against a neutral team.
Officially, by adding up the previous season's wins and losses of all 17 opponents and computing a winning percentage. Analytics sites instead average the opponents' current power ratings or market win totals, which describe the teams a schedule actually contains.
The record-based version, barely. The rating-based version, somewhat: a team facing opponents rated two points better on average than another team's slate projects for roughly one fewer win over a season. Neither is as important as the team's own quality.
It depends on which version is used and when it is measured. The site's team pages carry a rating-based schedule difficulty updated weekly from the current power ratings, and the remaining-schedule version during the season.
Category: Stats & Analytics. Part of the StickToTheModel NFL Encyclopedia.