Football is often described as a game of inches—where being off by even one can determine the difference between victory and defeat in a league with limited opportunities. The unique shape of the football itself introduces a degree of unpredictability, influencing everything from field position on punts to the way the ball spins on field‑goal attempts. Randomness also affects when points are scored, which can ultimately shape a team’s final win–loss record. Let’s face it, luck often becomes a deciding factor in a football season, and the best teams don’t always beat the worst ones. Take September 7th of last year, for example, when the Raiders defeated the Patriots 20–33. How does that happen?
As NFL analysts and sports bettors, we aim to develop reliable methods to predict future outcomes with greater accuracy. Wins and losses offer a glimpse, but they rarely provide the full picture or serve as strong indicators of future success. By identifying scientific approaches that connect past performance to future results, we deepen our understanding and improve our profitability. Today, we’ll explore the 2025 Pythagorean win‑total calculation as a predictive tool for the 2026 NFL season.
Can an NFL team outscore its opponents and still finish with a losing record? Or score fewer points than it allows yet end up with a winning record? Surprisingly, both scenarios occur almost every year. In 2025, for instance, the Baltimore Ravens scored 424 points and allowed 398, yet finished with a losing record of eight wins.
Teams often pile up points in blowouts but lose tightly contested games—a pattern that highlights the need to evaluate performance beyond surface‑level results. By analyzing data such as total points scored and total points allowed, we gain a clearer understanding of a team’s expected performance compared to its actual results. This approach, known as the Pythagorean win‑total method, helps uncover discrepancies and project realistic expectations for the following season.
According to the 2011 edition of the Football Outsiders Almanac, “From 1988 through 2004, 11 of 16 Super Bowls were won by the team leading the NFL in Pythagorean wins, while only seven were claimed by the team with the most actual victories.” The statistical significance of the Pythagorean wins theorem has held up ever since.
The Pythagorean theorem—expressed as a2+b2=c2—calculates the hypotenuse of a right triangle (c), or in our case, the expected relationship between points scored and points allowed. Imagine points scored as (a) and points allowed as (b). The value of (c) is derived from the squared relationship between the two. By understanding the ratio a2/c2 (where c2=a2+b2), we can estimate the percentage of games a team should win, then multiply that percentage by the total number of games. A tall, narrow triangle represents a team scoring far more points than it allows; a short, wide triangle represents the opposite.
You might wonder why we don’t simply use points scored divided by total points as the basis for predictions. The answer is simple: it doesn’t work. If a team scores 75% of the total points and allows 25%, would you expect them to win 75% of their games? Absolutely not. A team averaging 30 points per game while allowing just 10 would be expected to win far more. Consider the 2007 New England Patriots, who went 16–0. They scored 68.3% of the total points (36.8 per game on offense, 17.1 allowed on defense) and far exceeded 12 wins. Their Pythagorean win total—13.76—aligns much more closely with their dominance.
But there’s more to consider. The basic Pythagorean equation has limitations if left unadjusted. It tends to pull teams toward the middle, even when real‑world outcomes vary more dramatically. To account for these differences, adjustments must reflect the nuances of each sport—football included. Aaron Schatz of Football Outsiders suggests that each team’s exponent should vary based on its points scored.
According to Schatz, the formula for the exponent that works best in the NFL is:
1.5 × log((PF + PA) / G)
Much of the remaining randomness comes from inconsistent officiating, injuries, and pure luck. The average exponent across teams is 2.46.
One factor we can partially account for is turnovers. Every year, teams experience what’s known as turnover luck. Whether good or bad, turnovers directly affect scoring and game outcomes. A team in field‑goal range might fumble away three points, while a defense may be put in a terrible position after an offensive turnover deep in its own territory. Last season, the Bills had the best turnover margin at +24, while the Browns had the worst at –22. No surprise there.
Turnovers are widely considered to be worth about four points each, and the traditional view is that they’re mostly random. I disagree. Recent team histories show that turnovers are somewhat predictable. Before Tom Brady arrived in Tampa Bay, Jameis Winston consistently produced negative turnover margins for the Buccaneers. Indianapolis has also been reliably negative for years. On the other hand, some teams consistently protect the ball, while others field ball‑hawking defenses that force more turnovers than average. The Steelers are excellent at ball security, and the Ravens—dominant in their division—have posted positive turnover margins in each of the last five seasons.
Because of this, I’ve adjusted the turnover value from four points to two, awarding or penalizing each team one point on offense and one on defense based on its 2025 turnover ratio. Normalizing turnovers gives us a clearer picture of what each team’s expected wins should have been in a cleaner, more neutral season. In essence, teams with negative turnover margins receive a boost in expected wins, while teams with unusually high positive margins see their expected wins reduced to account for luck. Even though some teams consistently take care of the ball, outliers still exist. Fumble luck, in particular, is far less predictable than interceptions.
Please see our chart below for actual wins versus Pythagorean expected wins.
Using this diagram, we can make informed judgments about each team’s true strength—not only how well they score, but how well they defend. By comparing last season’s actual results with Pythagorean expectations, we can identify which teams overachieved and which underachieved in 2025. Here are the five biggest outliers on each side:
Over-achievers:
- Chicago Bears: Da Bears had 8.04 expected wins compared to their 11 actual wins. A -2.96 difference. They were certainly the luckiest team in the NFL for the 2025 season. A massive +22 turnover margin was the icing on the cake. This Bears team was the 4th quarter comeback kids. Can they make the playoffs or even win the division going from a 4th place schedule to a 1st place schedule? Well, that will mainly be up to the coaching and the quarterback play.
- Denver Broncos: Good ole Denver was here to clean up the misfortunes of the Chiefs. Well, until the playoffs that is. The Horses had had only 11.21 expected wins compared to their 14 actual wins last year. A -2.79 difference. Now with a first place schedule, and a healthier division, will this team win anywhere close to 14 games?
- Pittsburgh Steelers: Who is not sick of this slightly above average team? Steeler nation had expected wins of 8.12 compared to their 10 actual wins last year. A -1.88 difference. Aaron Rogers swoops in to achieve his normal regular season glory, only to crap out in the playoffs. The usual.
- LA Chargers: The Bolts had expected wins of 9.2 compared to their 11 actual wins last year. A -1.80 difference. Another team in the AFC West who benefited from an easier schedule due to the Mahomes injury and the Raiders just being the Raiders. Can Jim Harbaugh keep it boring again for the 2026 season?
- Jacksonville Jaguars: This London loving team had expected wins of 11.36 compared to their 13 actual wins last year. A -1.64 difference. This team was also quite fortunate with a +13 turnover margin before they choked away their chances at home in the playoffs. Thus far, Liem Cohen has proven to be a solid heard coach, so that is what we need to factor in for 2026.
Under-achievers:
- Kansas City Chiefs: The Chiefs had expected wins of 9.57 compared to their 6 actual wins last year. A +3.57 difference. Now clearly much of this was due to the bad injury to quarterback Patrick Mahomes, but even before that, we must remember from last year that this team won 15 games with a Pythagorean win total of 9.86 which made them the biggest regression candidate for 2025. The Chiefs under 11.5 wins was our best bet from last year’s article.
- NY Giants: The G-Men had expected wins of 7.10 compared to their utterly gross 4 actual wins from last season. A +3.10 difference. The Giants moved on from their coach Tony Daboll and hired a proven commodity in John Harbaugh. The Giants also have a quarterback with Jaxson Dart, and one hell of a running back in Cam Skattebo. As this division opens up, I think that the Giants are on the right path to the playoffs.
- Arizona Cardinals: The Red Birds had expected wins of 5.33 compared to their 3 actual wins from last season. A +2.33 difference. Lot’s of changes in this organization as they say goodbye to coach Rich Gannon and hello to new coach Mike LaFleur. I am not a big fan of the little brother perspective in general, but this team still needs a quarterback after they sent Kyler Murray and is Xbox over to Minnesota.
- Indianapolis Colts: The Ponies had expected wins of 9.93 compared to their 8 actual wins from last season. A +1.93 difference. The demise of this team after Daniel Jones’s injury was significant. Imagine starting the season 8-2 and then losing your last seven games to finish out the season. The good news is that this is a division that can be conquered for the 2026/2027 season.
- Washington Commanders: The Commie’s had expected wins of 6.7 compared to 5 actual wins. A +1.70 difference. From rags to riches and back to rags again was the story when rookie of the year quarterback Jayden Daniel’s dislocated his elbow. He tried to come back too early and re-injured it. This team however should have won a few more games with Marcus Mariota as they lost two games in overtime.
Best Bet: Giants over 7.5 wins +105
Some fans of the Pythagorean win total formula do not want turnovers involved for the argument that turnovers are more predictible. I do not agree with that assessment, but to try and keep everyone happy, I decided this year to insert a chart here below that shows the formulas without factoring in turnovers. Notice that the most fortunate team in the Chicago Bears who had a +22 turnover margin moves down to second most fortunate with a -1.84 difference as the Broncos are now in first In that category now up -2.95 wins. 
In conclusion, it should now be easier to see the randomness in football were some teams will have better records than they actually deserve, while some teams will have worse records than they actually should have achieved. Remember to take this information for what it’s worth. When predicting season wins, we also must factor in some of the less quantifiable information such as the clutch performances from quarterbacks like Josh Allen and Jalen Hurts along with the anti-clutch performances of Kirk Cousins and Geno Smith. Strength of schedule is another data point that is less quantifiable when it comes to actual points or season wins, but equally important when determining what these teams went through, not only in the previous season, but also in what they will be facing in the near future.
Pythagorean win totals are an important indicator of the future success or failure of NFL teams going into their 2026/2027 season, and it should take some precedence over last year’s actual win results. The Pythagorean win total will remain a great method for your predictive analysis for next season and beyond. Should you have any questions or want a list that can be copied and pasted, please tweet me @OBKiev. Have a great NFL season and best of luck in your research.
