Why 2025 NFC East standings are a bad predictor of 2026 standings

Every summer, NFL fans ask the same question: How good will my team be this year? The answers usually fall into one of two camps. One side points to free agency, the draft, and coaching changes. The other points to last year’s record.

Neither tells the whole story.

Long before analytics became mainstream, baseball writer Bill James proposed six indicators that tended to predict whether teams would improve or decline from one season to the next. The indicators were designed for baseball, not football, but the underlying ideas apply remarkably well to the NFL.

Some of these indicators are stronger than others. None should be treated as definitive. The NFL is an unusually noisy sport. Teams play only 17 games. Injuries swing seasons. One tipped interception can change a division race. That means even the best statistical indicators should be viewed as tendencies rather than predictions.

Our goal today isn’t to predict the 2026 standings. It’s to identify the forces quietly pushing teams toward improvement, or regression, before the season begins.

Let’s see what they say about the 2026 NFC East.

One of the strongest predictors of future performance isn’t a team’s record, it’s its point differential.

Wins matter, of course. But not all wins are equally convincing. A team that consistently outscored opponents by two touchdowns is usually more sustainable than a team that survived a string of one-score games.

That’s why analysts often start with point differential rather than the standings.

The Pythagorean Formula estimates how many games a team should have won based solely on points scored and points allowed. If a team wins significantly more games than expected, history suggests some regression is likely. If it wins fewer games than expected, improvement often follows.

Here’s how the NFC East stacks up entering 2026:

The biggest takeaway isn’t that the formula predicts exact records. It doesn’t. Rather, it identifies teams whose record may have overstated or understated their true level of play.

The Eagles outperformed their expected wins; the Cowboys were about as expected. The Commanders and Giants finished with fewer wins than their point differential deserved, and are likely to improve, even without major roster changes.

This is one of the most reliable indicators we have, because point differential tends to be much more stable from year to year than win-loss record.

One-score games have an outsized impact on NFL standings. The difference between a 12-win season and a 9-win season is often just a handful of plays spread across four or five close games. While elite teams may consistently put themselves in position to win late, the outcome of those games is surprisingly volatile from year to year.

Here’s the NFC East record in games decided by seven points or fewer:

Teams that go 8–2 in one-score games rarely repeat that success. Teams that finish 2–8 usually improve, even if the roster stays largely the same. That doesn’t mean close-game performance is meaningless. It means extreme close-game performance is difficult to sustain.

If a team’s record was built on an exceptional performance in close games, history suggests some of that success is likely to fade – and vice versa.

These two indicators are stronger together than they are separately. If a team outperformed its Pythagorean expectation and posted an exceptional record in close games, the case for regression becomes much stronger. If a team underperformed its expected wins and struggled in close games, the case for improvement becomes much stronger.

And that leads directly to the second question.

If some of last year’s results were inflated by close-game luck, how much of the overall season should we expect to repeat?

Every NFL fan believes last year’s record tells us something, especially if it’s a good record. And the record does tell us something. Just not as much as we tend to think.

Every NFL season produces a handful of surprises. A team comes seemingly out of nowhere to win 13 games. Another falls from the playoffs to the bottom of the division. Teams improve, coaches innovate, and players develop, but extreme outcomes also tend to include an unusually large dose of good or bad fortune. Once that fortune fades, records often drift back toward the middle.

That’s why the next two indicators focus less on what happened and more on how likely it is to happen again.

Regression to the mean is one of the most misunderstood concepts in sports. It isn’t a force of nature pulling every team toward .500. It’s a statistical observation that extreme outcomes are often followed by less extreme ones because the original result was influenced by factors that are difficult to repeat.

Think about a team that finishes 14-3. To reach that record, it probably needed a talented roster, good coaching, solid quarterback play, relatively good health, favorable turnover luck, and perhaps a couple of fortunate bounces in close games.

It’s entirely possible that every one of those things happens again. It’s just unlikely.

The same logic applies at the other end of the standings. A four-win team isn’t automatically destined to improve, but it doesn’t need everything to change in order to win six or seven games the following year. Simply returning to league-average health or turnover luck can make a meaningful difference.

The approximation for regression is intentionally simple. Start with 8.5 wins, then add or deduct one-quarter of a win for every game last year above or below .500.

The result isn’t a prediction of where each team will finish. It’s a reminder that last year’s record probably exaggerates both the highs and the lows.

Notice what’s happening: The best teams are pulled slightly back toward the middle. The worst teams are nudged upward. That isn’t pessimism or optimism. It’s probability.

Regression to the mean looks at where teams finished. The Plexiglass Principle looks at how they got there, and it’s quite simple. Teams that improve dramatically from one season to the next often take a step backward the following year. Teams coming off disastrous seasons frequently bounce back.

James called it the Plexiglass Principle because teams that rise quickly have a habit of bumping their heads on an invisible ceiling. The name is memorable. The explanation is much less mysterious.

In simple terms, teams that show a dramatic improvement (or decline) from one season to the next have a tendency to relapse (or bounce back) in the following season. The Cowboys of course have shown all the bouncing ability of a brick wall over the last two years with their succession of 7-win seasons. But the other NFC East teams did bounce around a lot.

The question here is which of the two most recent seasons (2024 & 2025) to consider the norm and which the aberration. We’ll consider 2025 the baseline, but an argument can be made for both the Eagles and Commanders that 2024 was the aberration, and the bounce back already happened in 2025.

Unlike Pythagorean wins or regression to the mean, the Plexiglass Principle doesn’t tell us why a team improved or declined. That’s where football knowledge still matters.

Did the improvement come from a new quarterback? A coaching change? An elite rookie class? Or was it driven primarily by unusually good injury luck and a favorable record in one-score games?

The indicator tells us to ask the question. It doesn’t answer it for us.

The first two indicators asked whether teams earned their record. These two ask whether that record is likely to be repeated. Sometimes the answer is yes. Dynasties exist for a reason. Elite quarterbacks continue to be elite. Great organizations stay well run. Sustainable success is absolutely possible.

But history suggests that most NFL seasons are less stable than they appear. A division champion isn’t automatically the favorite to repeat. A last-place team isn’t automatically doomed to finish there again.

The standings often tell us where teams finished. They don’t always tell us where they’re headed.

By now, a pattern should be emerging.

None of those ideas predicts the future. They simply tell us where history suggests we should be skeptical. But football isn’t played on spreadsheets.

Every season introduces variables that no statistical model can fully capture. Coaching changes reshape offenses overnight. Rookie classes outperform expectations. Quarterbacks develop, or don’t. Injuries derail contenders, while previously overlooked players suddenly become stars.

That’s why the final two indicators deserve a little less weight than the previous four.

They’re less predictive, but they help provide context that raw numbers can’t.

One of Bill James’ original indicators looked at how baseball teams performed after the All-Star break. The theory was straightforward: teams that finished the season playing better baseball often carried that momentum into the following year.

The NFL doesn’t offer the luxury of a 162-game schedule. With only 17 games, we’re working with a much smaller sample, which makes this indicator inherently less reliable. A hot month in December can reflect genuine improvement—or simply a favorable stretch of opponents.

Still, it’s worth asking whether teams played better – or worse – than their overall record suggests during the second half of the season.

Using the Pythagorean Formula over the last eight games of the season gives us one way of answering that question.

Given the limited sample size, this indicator deserves less weight than the previous indicators. But it can still offer clues about whether a team’s trajectory matched its final record.

Suppose a team like the Giants (and to a lesser degree, the Eagles) finished the year playing better than its record suggested. That doesn’t guarantee improvement, but it does strengthen the broader case.

Conversely, if teams like the Cowboys and Commanders overperformed their projection over the final eight games, they are more likely to regress downward.

Of all six indicators, this is probably the weakest. And team age across 22 starters and a 53-man roster contains so much noise it’s probably close to useless.

But one area where football history and our own understanding of football agree that age can be important is the offensive line. Because it is so critical for the success of an offense, any age-related decline of the five guys up front can have a significant impact on team success. Or in Jim Harbaugh’s words:

ESPN’s John Clayton explained in his “Theory of 150” that if the combined age of your starting offensive line exceeds 150 years (or a 30.0 average), you should expect a decline in performance. And older O-lines are probably more prone to injuries, are possibly facing a strong, age-related decline in performance, and they also cost a ton more than younger O-lines.

Here’s how all NFL teams compare in average offensive line age on opening day 2026, based on the Ourlads.com depth charts.

This is probably the weakest indicator in this framework, but ‘weakest’ doesn’t mean ‘useless.’

Individually, none of these indicators tells us very much. No single statistic can predict an NFL season. If one did, sportsbooks wouldn’t exist. But collectively, they begin to paint a picture.

Each of these six indicators captures a different aspect of team performance. Some are built on decades of statistical evidence. Others are little more than educated guesses. None should be viewed in isolation. The real value lies in the overlap. When several indicators point in the same direction, they begin to tell a more convincing story than any one of them could on its own.

Here’s how the six indicators stack up for the NFC East entering the 2026 season.

Notice that there isn’t a projected win total. That’s intentional. These indicators aren’t trying to tell us that Team X will finish 11-6 or Team Y will go 8-9. They’re answering a different question.

Which teams are most likely to exceed expectations, and which teams may struggle to repeat last year’s success?

That’s a much more realistic use of historical data.

If the indicators collectively point toward another season hovering around the middle of the pack, they aren’t saying Dallas is doomed to finish with another seven wins. They’re saying something subtler.

Standing still is remarkably easy in the NFL.

Moving out of the middle usually requires something meaningful to change. Sometimes these changes are obvious:

Sometimes it’s less obvious.

Those are football questions. The statistics can’t answer them. What they can tell us is whether those changes need to happen in the first place. And for teams that repeatedly find themselves clustered around .500, the answer is usually yes.

Every preseason is built on optimism. Every fan base can point to reasons why this year will be different. Sometimes they’re right.

Sometimes the biggest difference isn’t a coaching hire, a draft class, or a blockbuster free-agent signing. Sometimes it’s simply the way probability works. That’s what makes Bill James’ framework so useful, even after all these years. Not because it predicts the future. Because it reminds us to question the past.

The standings tell us who won. They don’t always tell us who played the best football. They tell us where teams finished. They don’t always tell us where they’re headed.