We can look at Pythagorean expectation, which predicts the expected number of wins based on points scored vs. points allowed at this point in the season. And then teams that are winning more games than they “should” based on points scored vs allowed are “worse than their record”, and conversely teams that are winning fewer games than they should are “better than their record.”

By 9 or 10 games played in the expansion era (since 2002), 62% of teams are within 1 win of their expected wins. About 87% are within 2 wins. This season, interestingly, 100% of teams are within 2 wins of expected, and 85% are within 1 win.

There is 1 team who is 2 wins below expected (better than their record):

• Buffalo Bills. Expected: 7-3. Actual: 5-5.

There are 4 team who are 2 wins above expected (worse than their record):

• Philadelphia Eagles. Expected: 6-3. Actual: 8-1.
• Detroit Lions. Expected: 5-4. Actual: 7-2
• Seattle Seahawks. Expected: 4-5. Actual: 6-3.
• Pittsburgh Steelers. Expected: 4-5. Actual: 6-3.

Now personally I’d argue that no team is really better or worse than their record, because a team that is getting significantly more wins than expected based on points probably has the benefit of skillful players and coaches who can figure out how to keep games close and exploit advantages to squeak out victories (e.g., the Steelers). And teams that are below expected conversely underperform in such situations by habitually making costly errors in close games (Bills). Also, it can be skewed by big blowout wins or losses which in some cases could be isolated “burn the tape” outlier games that don’t really reflect a team’s average performance (Lions and Seahawks vs. Ravens). So make of this what you will.