Week-one in the NFL: overreact, underreact, or just react?

Jason Pauley
4 min readSep 9, 2022


by Jason Pauley

What happens to teams the remainder of the season after a loss or win in week one? I’m convinced, as most of you probably are, that there is nothing magical about week-one. It’s 5.9% of the season, just like week two or week three. And my hunch is the data below would look very similar for ANY week (excluding the last couple of weeks late in the season when teams are sitting players and positions in the standings are already settled). But week-one is the next week in the season, so we can look at this from the lens of week-one.

A team that wins any given week is more likely to be better than a team that losses on any given week, so none of this should be surprising. It’s more about understanding to what magnitude (big or small) a team’s final record falls in line with our expectations for 1–0 or 0–1 teams.

I looked at data from 2000 to 2021 to see what happens after a week-one lass or win.

Note: all data is prorated over a 17-game season. Playoff rates cross eras. More teams make the playoffs now, so a team is a little more likely to make the playoffs today than the data suggests.


Historically teams win the equivalent of an additional 8.6 games the rest of the season (prorated to the 17-game season) after a week-one win, and end up with 9.6 wins and 7.4 losses over 17 games. Teams with a 0–1 start will win 7.4 games for the remainder of the season and end up with 7.4 wins and 9.4 losses.

The success rate for making the playoffs is 2X greater for a 1–0 start compared to 0–1, and the rate to make the Super Bowl has been 3X greater with a 1–0 start vs a 0–1.

My recommendation is to React.

Point Differential

Point differential doesn’t seem to follow a pattern. In other words, there isn’t a consistent decrease in remainder of the season winning percentage if a team’s week-one loss point differential increases. When a week-one win point differential increases, there is an increase for the first three categories, but then for the really big wins (N=22) the winning percentage drops significantly. I also looked at the R-Squared between week-one point differential and the winning % for the remainder of the season and it was a paltry .043. Not much pattern to any of this.

A fun example to go along with the data: The 2003 Patriots got the sh*t kicked out of them in week one of 2003 when the Bills beat them 31–0. The Patriots went 14–2 and won the Super Bowl, the Bills ended up going 6–10. There were probably a lot of overreactions to that game a score at the time.

My recommendation for week-one point differential is to Underreact.

Home and Away

Overall, the table at the top shows what happens regardless of home/away situation. Does a home loss matter more? Does a road win matter more? It turns out, in terms of the remainder of the season, as week-one loss is a loss and hasn’t mattered where the loss happened.

Lose at home…7.3 wins the rest of the year; lose on the road 7.4 wins. Playoff rate is 24.7% and 24.6% for those two scenarios, virtually the same.

But look at the road wins. Teams winning on the road on in week one win 8.8 games (prorated to 17-gm season) for a .551 win %, compared to teams winning at home (8.5, .531); ending with a 9.8 win and 7.2 loss season including the win in week one. The payoff rate increases from 49% to 56% when starting off with a road win vs a home win. 10.5% of teams beginning the season with a road win, went to the Super Bowl.

My recommendation for week-one results relative to being home or away is to Underreact for losses, But we are fans, and if we aren’t overreacting about something, then what’s the point? My recommendation for week-one road wins is to overreact. Celebrate that week-one road win like you just won the Super Bowl because that’s probably what is going to happen anyways, right?



Jason Pauley

Passionate about Analytics (Football, Sports, Marketing, Sales, Demographics)