Why Hot ATS Streaks Fade

The test

Take every team, every season from 2015 to 2024, and put its cover rate next to its cover rate the following season. That gives 320 pairs. If covering were a lasting trait of a team, teams that covered a lot one year would tend to cover a lot the next, and the correlation between the two columns would be clearly positive.

It is −0.02. A correlation runs from −1 to 1, and zero means knowing one number tells you nothing about the other. A team that covered 65% last year is, as far as this history can tell, no more likely to cover this year than one that covered 35%.

Why it happens

A team that covers a lot is a team the market underrated. Underrating is not stable: every cover is public, and the lines are set by people with a great deal riding on learning from them. Once a team is known to be better than its reputation, the spread moves to meet it, and the easy covers stop.

Luck plays a bigger part than it looks. Over a 17-game season a coin flip against the spread lands anywhere from about 26% to 74% (two standard deviations either way), so a “hot” ATS season is often just the upper half of chance. Across 2015 to 2025, the 32 teams' cover rates spread out by 3.7 percentage points, exactly what chance produces (see what an ATS record means).

Over a few weeks, it runs backward

Over a handful of games the pattern does not just vanish; it tilts the other way. The NFL Spread Dial includes each team's cover rate over its last six lined games as one of its factors, and the fitted weight comes out negative: in the September 2026 fit it was −1.50 points per unit of cover rate. In plain terms, when the home team has covered all six of its recent games and the road team three of six, the dial moves its estimate about three-quarters of a point against the hot team.

That is a small number and it is easy to explain away. The dial shows it anyway, with its sign, rather than quietly dropping a factor that did not say what intuition expects. You can turn that dial to 0× to ignore recent covers entirely and see what it does to the back-test.

What does carry over

Team quality persists from week to week and, less strongly, from season to season. That is what a power rating like the Spread Dial's Elo measures. But quality is also exactly what the line is built on, so a good team's next game is priced as a good team's game. The part of a team's results that the line did not expect is the part that does not repeat.

Every team page under NFL ATS Records shows its season-by-season cover rates side by side, so you can see the zigzag for yourself.