The Super Bowl hangover is one of the most interesting supposed phenomena in the NFL. According to this theory, any team that makes the Super Bowl should be expected to drastically fall off in the subsequent season. After the Kansas City Chiefs lost the first game of the season in surprisingly convincing fashion, this theory has once again reared its head among football media circles and fans.
Let’s assume for a moment that this does exist. How can we prove – or disprove – the existence of the Super Bowl hangover? In today’s piece, I’m going to examine the past performances of prior Super Bowl teams as well as how they performed in their subsequent seasons.
Brief Methodology Notes
Due to the inherent noisiness and binary nature of win-loss totals across non-regularized season sizes, I’ll be using FTN’s historical end-of-season DVOA as a stand-in.
For those of you who are not familiar with DVOA, it is one of the most respected, if not the arguably the most important, football metrics in the world. DVOA measures a team’s performance on the average play in all three phases of the game (offense, defense, special teams), adjusted for opponent quality, down, distance, field position, and league average results on each play.
What Does The History of Super Bowl Losers Suggest?
To ensure I was looking at a relevant sample size, I only looked at teams that lost the Super Bowl in the 21st century and their follow-up seasons. This excludes the 2024 Chiefs, yet includes every Super Bowl loser from the 2000 New York Giants to the 2023 San Francisco 49ers. I then tracked each team’s DVOA during their respective seasons, as well as their follow-up seasons, and calculated the difference in DVOA from year to year.
| Year | SB Loser | DVOA (RS + PS) | DVOA After (RS + PS) | DVOADiff |
| 2023 | SF | 36.3% | 6.7% | -29.6% |
| 2022 | PHI | 27.8% | 1.1% | -26.7% |
| 2021 | CIN | 3.9% | 20.7% | 16.8% |
| 2020 | KC | 22.5% | 20.4% | -2.1% |
| 2019 | SF | 35.9% | 2.6% | -33.3% |
| 2018 | LAR | 23.9% | 8.4% | -15.5% |
| 2017 | NE | 22.3% | 13.3% | -9.0% |
| 2016 | ATL | 31.0% | 8.4% | -22.6% |
| 2015 | CAR | 23.7% | -4.9% | -28.6% |
| 2014 | SEA | 29.6% | 35.4% | 5.8% |
| 2013 | DEN | 33.5% | 32.6% | -0.9% |
| 2012 | SF | 29.2% | 16.6% | -12.6% |
| 2011 | NE | 24.0% | 35.3% | 11.3% |
| 2010 | PIT | 31.4% | 24.0% | -7.4% |
| 2009 | IND | 4.4% | -36.4% | -40.8% |
| 2008 | ARI | -7.1% | 16.3% | 23.4% |
| 2007 | NE | 52.3% | 16.4% | -35.9% |
| 2006 | CHI | 24.6% | -4.0% | -28.6% |
| 2005 | SEA | 26.4% | -12.1% | -38.5% |
| 2004 | PHI | 25.6% | -3.0% | -28.6% |
| 2003 | CAR | -1.1% | -4.8% | -3.7% |
| 2002 | OAK | 27.2% | -15.9% | -43.1% |
| 2001 | STL | 31.2% | -2.6% | -33.8% |
| 2000 | NYG | 3.6% | -10.4% | -14.0% |
In total, we have 24 teams that reached the Super Bowl and lost. On average, the “21st Century Super Bowl Loser” had a 23.4% DVOA in their initial season, then suffering a dropoff of -16.6% DVOA in their following season to an average following DVOA of 6.8%. On the surface, these results seem to prove that there does exist a dropoff – only four of the 24 teams improved by DVOA in the following season.
With that said, the standard deviation in DVOA shifts is around 18.9%, making for a heavily volatile sample. On a qualitative basis, the 2017 New England Patriots and 2020 Kansas City Chiefs each were followed up by Super Bowl winning iterations of their respective franchise, but on a quantitative basis, the Bengals actually improved their season-long performance following their initial Super Bowl appearance, and the same could be said about the 2008 Arizona Cardinals.
What Does The History of Super Bowl Winners Suggest?
Out of curiosity, I decided to look at the history of Super Bowl winners to see if there was any significant difference between a Super Bowl hangover between winning teams and losing teams in their following seasons. However, the results were fairly surprising.
| Year | SB Winner | DVOA (RS + PS) | DVOA After (RS + PS) | DVOADiff |
| 2023 | KC | 17.9% | 7.3% | -10.6% |
| 2022 | KC | 26.8% | 17.9% | -8.9% |
| 2021 | LAR | 24.9% | -13.3% | -38.2% |
| 2020 | TB | 26.8% | 23.9% | -2.9% |
| 2019 | KC | 30.5% | 22.5% | -8.0% |
| 2018 | NE | 13.3% | 24.2% | 10.9% |
| 2017 | PHI | 22.5% | 3.7% | -18.8% |
| 2016 | NE | 25.4% | 24.0% | -1.4% |
| 2015 | DEN | 17.1% | 1.2% | -15.9% |
| 2014 | NE | 21.6% | 23.4% | 1.8% |
| 2013 | SEA | 35.8% | 29.6% | -6.2% |
| 2012 | BAL | 8.8% | -9.5% | -18.3% |
| 2011 | NYG | 5.9% | 10.8% | 4.9% |
| 2010 | GB | 19.7% | 24.6% | 4.9% |
| 2009 | NO | 24.2% | 11.9% | -12.3% |
| 2008 | PIT | 24.9% | 17.1% | -7.8% |
| 2007 | NYG | 4.5% | 27.3% | 22.8% |
| 2006 | IND | 18.3% | 29.5% | 11.2% |
| 2005 | PIT | 23.6% | 9.0% | -14.6% |
| 2004 | NE | 35.9% | 8.4% | -27.5% |
| 2003 | NE | 20.2% | 35.9% | 15.7% |
| 2002 | TB | 30.7% | 7.6% | -23.1% |
| 2001 | NE | 8.3% | 12.8% | 4.5% |
| 2000 | BAL | 21.1% | 6.7% | -14.4% |
Within a roughly equal sample size of Super Bowl winning teams, there are quite a few interesting differences. For starters, the average DVOA of a Super Bowl winning team was actually lower (though not necessarily significant) at 21.20%. However, where the data is significantly different is the dropoff: with Super Bowl champions having a net -6.34% shift in their DVOA from their winning season to their follow-up season to 14.85%. This is essentially the difference between last year’s Chiefs and last year’s Indianapolis Colts. Noticeably, there is a difference in standard deviation for DVOA shifts from Super Bowl to post-Super Bowl seasons too (14.36%).
To double check whether or not this was statistically significant, I conducted an unpaired t-test comparing the group means, standard deviations, and sample sizes of both the Super Bowl winners and Super Bowl losers. The two-tailed P value equaled 0.0396, meaning that there’s roughly just under a four percent chance that any difference in DVOA shifts between the two groups would be entirely random. As it stands, it does seem like recent Super Bowl losers do suffer statistically significant dropoffs from Super Bowl winners.
However, keeping this in mind, I think that there is an even better way to describe the Super Bowl hangover, regardless of whether it impacts losers or winners.
Is This Simply Just Regression to the Mean?
To test this, I expanded the dataset to include every NFL team since 2000 and modeled how team strength carries over year to year. I ran a regression of each team’s DVOA in season N against its DVOA in season N+1.
Though volatile and simplified,, the results show a clear pattern: great teams tend to decline, bad teams tend to improve, and average teams hover near league average. In other words, what we call a ‘Super Bowl hangover’ may be a good old fashioned statistical reality in any volatile competitive field.
To make the results more specific, I bucketed every team since 2000 into five groups: GOAT (≥25%), Great (10–24%), Average (–10% to +10%), Bad (–10% to –19%), and WOAT (≤–20%). I then examined their typical DVOAs during their seasons and their follow up seasons.
GOAT teams fell from +31% to +15% DVOA the next year, similar to our Super Bowl losers. Great teams declined from +17% to +10% – roughly similar shifts to our Super Bowl champions. Meanwhile, Bad and WOAT teams improved by 8 to 16 percentage points on average, as the middle barely budged. No matter the direction, the more extreme the performance, the stronger the pull back toward the middle.
Keep in mind that together, Super Bowl winners and losers made up 48 instances within our total data set of teams with follow-up seasons (766 teams). Together, the Super Bowl teams had an average DVOA of 22.3% in their Super Bowl season, while their follow-up season had an average DVOA of 10.8%, making for a -11.5% DVOA shift. All the 227 teams that had a Great or higher (GOAT) classification had a shift of -10.0% (more specifically 9.97). Using a Z-test to calculate the p-value of our sample against the entire population of “great and GOAT teams”, I calculated a 0.57 p-value. In other words, there is roughly a 57% chance that any difference between the Super Bowl hangover and the “great and GOAT team hangover” (“the typical Super Bowl contender regardless of appearing in the Super Bowl or not”) is random.
Conclusion
While Super Bowl teams typically regress, the real question is whether or not they regress in a manner that’s truly unique compared to the rest of the league. The truth is that making the Super Bowl is not the only thing that happens before a decline. Roster changes, injury luck, strength of schedule, and the NFL draft are four factors among several others that influence whether or not a team takes a step back.
On one hand, there’s reason to expect Super Bowl losers to suffer a little bit more than winners. However, I don’t think that losing in the Super Bowl is somehow worse than losing in the conference championship game or anywhere earlier in the playoffs. Extreme performances in any direction are almost always followed by a return toward the pack. It just happens that making the Super Bowl happens to be an extreme performance and something largely composed of extremely good teams.
Based on my research, I do not think that the Super Bowl hangover meaningfully exists. If the 2025 Chiefs lose another game or struggle for the rest of the season, we’ll have to find a different – or simply more detailed – explanation.