NFL regression candidates: Patriots and Bears will have a tough time repeating in their divisions
It doesn’t seem like only 16 months ago that the Washington Commanders were the talk of the NFL. The Commanders, with then-rookie Jayden Daniels , upset the No. 1 seed Detroit Lions in the playoffs and then lost in the NFC championship game to a fantastic Super Bowl champion Philadelphia Eagles team. When a magical season like that happens, the natural thought is to believe a team has arrived and the next step will be forward. That’s not always the case. It certainly wasn’t for the 2025 Commanders. They crashed back to earth as the roster got old in a hurry, some luck in close games ran out and injuries hit. They went 5-12, a huge disappointment for fans who believed their big season was the start of something big. Play 2026 Soccer Pick 'Em with FOX One and make your picks for the world's biggest soccer tournament Regression is a strong force, and it’s bound to hit some of the breakout teams from last season. Here are the top five candidates to regress back to the mean this season, with the biggest reason for the pessimism: There’s one big factor here that could negate all of the regression factors working against the Bears, and it’s Caleb Williams . Williams still had accuracy issues last season, but showed significant growth and it’s reasonable to think there’s a big leap coming. If he was an MVP candidate in his third season, it wouldn’t be a big surprise. But plenty of regression indicators are not kind to the Bears. Start with turnover differential. The Bears were +22 last season, which was five better than any other team and nine better than third place. That’s likely unsustainable. There’s also the difference between the Bears’ 11-6 record and their Pythagorean win total, which calculates what a team’s record should have been based on their point differential. The Bears, who had a +26 point differential, had a three-win gap between their adjusted Pythagorean win total and actual win total according to TheOddsBreakers.com , the most fortunate mark in the NFL. Eight wins in one-score games, including one in the playoffs, helped. Also, the Bears go from the seventh easiest schedule in the NFL ( calculated by DVOA ) to the sixth toughest this season ( calculated by SharpFootball.com , using oddsmakers’ win totals). There are other metrics that say the Bears won’t repeat as an 11-win team. Williams’ improvement could even that out, but he might have to improve a lot to overcome some impending regression. We all know about last season’s strength of schedule for the Patriots, which was one of the easiest in recent NFL history. And if that was the only thing working against the Patriots, who went 14-3 last season, it would be significant but maybe not too much to overcome. That’s not the only reason though. The Patriots were the luckiest team in the NFL last season when it came to adjusted games lost due to injury, a metric from FTN Fantasy’s Aaron Schatz. That was a massive advantage and very hard to repeat. Seven wins in games decided by seven points or less is also going to be hard to repeat. New England had a great season and has a young core, with a quarerback in Drake Maye who nearly won NFL MVP and should still be improving. But a pullback is practically inevitable, it’s just a matter of how many fewer wins the Patriots get. Picking a bunch of teams that went 15-2 or 14-3 last season to regress and then doing a victory lap when they don’t win 14 games again is very low-hanging fruit. Of course a 14-3 team like the Broncos are very likely to win fewer games, what other way could they realistically go? But it’s also reality that the Broncos have regression coming. The Broncos won 14 games last season and went 11-2 in games decided by eight points or less. They were 6-2 in games decided by three points or less. Having a fantastic defense and a great coach in Sean Payton gives Denver an edge in close games, but a record that lopsided in close games is not repeatable. We saw with the 2025 Chiefs that a record far [... truncated ...]
