This post first ran on CrystlQuant.
On the second Sunday of November, the road team won almost everywhere. The Giants won 19-11 in Philadelphia. The Packers scored 45 in Foxborough and beat the Patriots by 18. The Texans put 44 points on the Chargers in Los Angeles, and the Bengals beat the Falcons 34-17 in Madrid, in a game that counted as an Atlanta home game. By the time the Bills finished off the Vikings 23-6 on Monday night, road teams had won 10 of the 15 games in Week 9.
One strange week would be a footnote. This one tipped the whole season over. Through November 9, 2026, NFL home teams have won 66 regular-season games and lost 68. The one tie was the Giants and Commanders at 27-27 on October 11. Road teams are ahead.
That is not supposed to happen. Our tables hold every NFL regular-season game since September 12, 1999, and from 1999 to 2018 home teams won 2,902 of 5,073 of them: 57.2%. No season in those twenty years finished below 52.8%. In 27 full seasons, road teams have won more games than home teams exactly once.
The season with nobody in the seats
That once was 2020, the year the experiment ran itself. Because of the pandemic, most stadiums were empty or close to it all season, and home teams finished 127-128-1, the only losing home record of any full season since 1999.
Here is the strange part. After 135 games in 2020, home teams were 66-68 with one tie. After 135 games in 2026, they are 66-68 with one tie. Same wins, same losses, same single tie. One of those seasons had cardboard cutouts in the stands. This one has crowds, and the count is identical.
2020 is the best evidence there is that a big share of home-field advantage was the crowd: the noise that makes a road offense go to a silent count, and the pressure that may lean on the officials. Take the fans out and the advantage almost disappeared. So why has it gone missing in a season when the fans are back?
It was shrinking before anybody noticed
Because 2026 is not a bolt from the blue. It is the latest step in a long slide. From 1999 to 2018 hosts won 57.2% of games. From 2019 to 2025 they won 992 of 1,871, or 53.0%. Even 2019, the last season before the pandemic, came in at 51.6%, and 2021, the first full season with crowds back, at 51.5%.
The market noticed. Measured in points, the closing price gave the home team an average head start of 2.4 points a game from 1999 to 2016. From 2021 to 2025 that head start was 1.6 points, and this season it is 1.7. Playing at home is still worth something, but what used to be close to two and a half points is now closer to one and a half.
Why it shrank is an argument, not a count. Teams travel and recover far better than they did in 1999, with charter flights, sleep plans and staff whose whole job is getting players fresh for Sunday. Offenses practice for noise all week. Replay review fixes more of the calls a loud building might once have bent. None of those can be counted in a box score, but all of them push the same way.
Seven home games that were not
Part of this season's number is in the schedule. Seven 2026 games so far have been played outside the United States: in Melbourne, Rio de Janeiro, three times in London, in Paris and in Madrid. The team listed as the host won one of them, the Jaguars' 19-17 over the Eagles at Tottenham on October 11. The visitors won the other six.
Take those seven out and hosts are 65-62-1 at their own stadiums. Still ahead, but barely, and nowhere near the 57.2% of 1999 to 2018.
A lead at this point of the season has also been lost before. After 135 games of 2021, road teams were ahead 70-65. Home teams then won 75 of the remaining 137 games, lost 61 and tied one, and finished the year 140-131-1. Half a season is a long time in the NFL.
What CrystlQuant's numbers say
CrystlQuant publishes two win chances for every NFL game before kickoff. They say nobody beats the market. One model never sees the market price. The other fixes it.
Over the 134 games of 2026 that had a winner, the closing price gave the home team about 55% on average, which works out to about 74 home wins. Home teams won 66. The CQ blind model, which builds its numbers without ever seeing the price, leaned on home field even harder, at more than 57% on average. CrystlQuant itself went the other way. Across the 104 games it has published since September 22, CrystlQuant has the home team at 51.2% on average, against the market's 54.1% in the same games. The hosts won 51 of those 104.
Here is our take. Home field in the NFL is now a small thing, about a point and a half, and this season it has been smaller still. A forecast that still gives the home side an old-fashioned bump is paying for something the league stopped delivering years ago, and the one number in our three that shaved home field the most is the one that has been closest to the truth this season. Whether hosts climb back above .500, the way they did in 2021, is for the second half to settle. Either way, the days when playing at home won 57 games in 100 are over. Every NFL game, with CrystlQuant's number beside the market's before kickoff and graded after, is on CrystlQuant's NFL page.
Every game this week, with our number and the market's side by side, is free at crystlquant.com.
Top comments (0)