New England Doesn't Need to Be Good to Be Useful

Fantasy football has trained us to hunt good offenses.

Find the touchdowns. Follow the points. Draft the quarterback. Draft his receiver. Maybe draft the running back too. By September, you own four players from the same team and have accidentally become emotionally invested in a third-and-7 from Foxborough.

The 2025 Patriots are exactly how this happens.

New England went 14–3, won the AFC East and reached Super Bowl LX one year after going 4–13. Mike Vrabel's first season did not produce a gentle rebuild. It produced one of those seasons that makes the previous three years look like an administrative error.

And now AutoDraftYears expects the offense to get worse.

Quite a bit worse.

The Patriots' offensive scoring input falls from 26.5 points per game in 2025 to 22.8 in 2026.

So the question is not whether New England was good last year.

It very obviously was.

The question is:

How much of 2025 should fantasy managers be willing to pay for again?

2025: Everything Changed at Once

It is difficult to overstate how different New England looked.

The Patriots had won eight total games across 2023 and 2024. Then Vrabel arrived, Josh McDaniels returned to run the offense, and New England won 14 games.

Most importantly for fantasy, Drake Maye stopped looking like a promising young quarterback and started looking like a problem everyone else in the AFC needed to discuss in a meeting.

Maye completed 72.0% of his passes for 4,394 yards, 31 touchdowns and eight interceptions, while adding 450 rushing yards and four more scores. His 72.0% completion rate, 8.9 yards per attempt and 113.5 passer rating all led the NFL.

That is not normal second-year development.

That is skipping several chapters.

Stefon Diggs also answered a fairly important question after coming back from his 2024 knee injury. He caught 85 passes for 1,013 yards and four touchdowns, including five 100-yard games. He has since signed with Washington, so that production leaves with him.

Rookie TreVeyon Henderson added 911 rushing yards, and the Patriots suddenly had something they had lacked for years:

Players fantasy managers actively wanted.

Officially, New England averaged 28.8 total points per game, second in the NFL. AutoDraftYears' offensive scoring measure strips out non-offensive scoring and puts the offense at 26.5, eighth.

Either way, the transformation was real.

So was the ending.

New England won three playoff games but scored only 10 points in the AFC Championship and then spent its first eight Super Bowl possessions punting before losing 29–13 to Seattle.

The season was spectacular.

It was not flawless.

And there is another piece worth remembering before we automatically extrapolate 14–3 into the future: Pro Football Reference rated New England's 2025 schedule at -4.49, one of the easiest in football.

This is roughly where AutoDraftYears enters the room carrying a bucket of cold water.

The Model Is Betting on Gravity

The model projects New England to fall from 26.5 offensive points per game to 22.8.

Not back to the horror movie that was 2023.

Just back toward normal.

This is where it gets weird.

New England is projected to score 3.7 fewer points per game while throwing substantially more often.

Fantasy football managers have now leaned forward in their chairs.

The positional environments are less encouraging:

Running back is the only environment that grades particularly well.

Receiver ranks 25th.

Quarterback ranks 23rd.

This is not the model waving a Patriots flag.

Comparable offenses produced 2.7 fantasy starters on average, almost identical to the 2.6 league-wide average across all 320 team-seasons in the sample.

And the team-level number is blunter than I expected.

The market ranks New England's fantasy roster 10th.

The model ranks it 24th.

Fourteen places, against a threshold of eight.

So this isn't a rounding error. The Patriots are one of the few rosters the model thinks the market is actively overpaying for — which is the 14–3 hangover expressed as a single number.

Which makes the individual player prices much more interesting.

You Do Not Have to Draft the Offense

Market and Model are ranks within the position. Edge is the model's position rank against the market's, after removing the drift that affects every player at that price — so it reads as “the model likes him this many places more than it likes the typical player who costs this much.” A call needs the edge to clear the position's bar (QB 5, RB 15, WR 11, TE 5 places), set at two-thirds of that position's own spread.

Two names matter.

One in each direction.

Drake Maye Is Expensive. The Model Still Says Pay.

Maye goes around pick 56 as QB3.

The model also ranks him QB3.

That sounds suspiciously like agreement until the adjusted edge enters the conversation.

After accounting for what typically happens to quarterbacks at this price, Maye still produces a +9 positional edge, clearing the model's QB buy threshold of seven.

The model projects 20 PPG with Low risk.

More importantly, Maye just gave us evidence that he can create an elite fantasy outcome without requiring New England to become an elite passing environment.

He led the NFL in completion percentage, yards per attempt and passer rating last season.

I am comfortable paying QB3.

Not because I expect another 14-win Patriots season.

Because I don't need one.

A.J. Brown Is the Price That Doesn't Work

New England replaced Diggs with a bigger name.

A.J. Brown is a Patriot now, and that is the kind of transaction that makes a fantasy manager sit up straight before checking anything else.

Then you check.

ADP 22. WR8.

The model has him WR28.

Adjusted edge: -19.

The receiver fade threshold is -11, so this one clears it without needing a discussion.

The talent isn't the argument. Brown has spent most of a decade as one of the best receivers in football, and his 77 role score says the model expects him to be used like it.

The argument is the building he just walked into.

New England's receiver environment ranks 25th. The offense is projected to lose 3.7 points per game. And Maye's extraordinary season did not come from feeding one dominant target — it came from a 72% completion rate distributed across the roster.

WR8 asks you to pay for the version of A.J. Brown who had Philadelphia's offense around him.

The model projects 12 PPG for the version who has this one.

Henderson Is the Opposite Problem

I like TreVeyon Henderson.

The model likes New England's running-back environment.

Unfortunately, everyone else has apparently also discovered these facts.

Henderson costs RB24.

The model ranks him RB22.

Adjusted edge: +4.

Fair.

Fantasy football desperately wants every player we like to also be undervalued. Sometimes the cashier knows what the item costs.

What Should Fantasy Managers Do?

Do not draft New England as though 2025 established a new permanent offensive baseline.

The Patriots went 14–3. Maye played at an extraordinary level. They reached the Super Bowl. Almost everything that could improve from 2024 did.

The model expects some gravity.

I do too.

But team regression is not the same thing as player regression.

That distinction matters.

Maye is still worth QB3 prices.

A.J. Brown is the one I would let somebody else buy.

Henderson is fine at cost.

Stevenson, Doubs and Henry do not give us enough reason to force anything.

And that's really the lesson here.

Fantasy managers spend enormous amounts of time deciding whether an offense is “good” or “bad,” as though we're reviewing restaurants on Yelp.

Sometimes that is too crude.

You can think the restaurant is about to lose a Michelin star and still order the steak.

New England can regress.

Drake Maye can still be excellent.

A.J. Brown can still be badly overpriced.

All three things are allowed to happen.

Sources: Auto Draft Years FF Model workbook — Team Data By Year, Team Stat Ranking, 2026 Team Context; Auto Draft Years 2026 consensus board (16 authoritative market sources); Player outcomes 2016–2025, Pro Football Reference via the project dataset. Team environment grades, forecasts and player model ranks are proprietary. Every figure on this page is generated from the dataset at build time; none is typed in by hand.