Brian Flores' Minnesota defense finished 2025 as one of the two or three best pass defenses in the NFL by almost any measure. It also gave up more per-play damage on throws of 20+ air yards than any other team in the league. Both of those things are true, and they're true for the same reason.

Table of Contents

1. The top-line résumé
2. The single weirdest split in the league
3. What Flores changed: pressure up, air yards down
4. The 2024-to-2025 deep-ball trade
5. Why this is a feature, not a leak

The top-line résumé

Minnesota's 2025 defense finished 3rd in the NFL in overall EPA/play allowed (-0.101) and 2nd against the pass (-0.148 pass EPA/play), trailing only Houston. Only three teams held opponents below -0.10 EPA/play on all snaps; the Vikings were one of them.

That's a serious jump from 2024, when the same franchise allowed -0.081 EPA/play overall and just -0.043 on pass plays. The pass-defense number more than tripled in the right direction in a single offseason.

The single weirdest split in the league

Now break the pass defense apart by target depth. Minnesota's 2025 rankings across the four standard depth tiers are, in order:

The Vikings were the best intermediate-pass defense in football and the single worst deep-pass defense in football, in the same season, on the same unit. On deep throws they gave up 19 completions in 39 attempts (48.7%) for 622 yards and 6 touchdowns against just 1 interception. +1.083 EPA/play means every deep target dropped in Minnesota's zip code was worth more than a point of expected scoring — a catastrophic per-play rate.

What Flores changed: pressure up, air yards down

The mechanism is the pass rush. Flores' scheme dialed the pressure up and got quarterbacks off their spot faster:

Sack rate climbed nearly three points. Nearly one in every five dropbacks against Minnesota ended with a hit on the quarterback. And the average pass thrown at the Vikings dropped almost a full yard shorter in the air — quarterbacks were checking down, dumping off, or eating a rush instead of holding for the shot.

You can see that squeeze show up as fewer opportunities down the field. Minnesota allowed just 39 deep attempts on the season. The next-lowest defense (New York Jets) allowed 51. The league average was 67. The Vikings faced 42% fewer deep shots than an average defense.

By rate (deep attempts as a share of total pass plays faced), Minnesota's 8.7% was the second-lowest in the league behind Pittsburgh's 8.1% — so this isn't just "they played fewer pass plays overall." Opponents genuinely threw deep against the Vikings less often than against almost anyone else.

The 2024-to-2025 deep-ball trade

The deep-ball splits from 2024 to 2025 tell a story about volume and about outcomes:

In 2024, Minnesota's defense picked off deep passes ten times — one of the highest deep-INT totals in the league — while allowing +0.324 EPA/play on the tier. The 2024 formula was: opponents throw deep on us 83 times, and about once every 8 attempts we take the ball away.

In 2025, opponents got 39 chances. Minnesota intercepted exactly one of them. Every shot that connected went uncontested, and none of the ones that fell incomplete got flipped into a takeaway. When the plays are that rare, one uncontested 60-yard bomb per game reshapes the per-play average by itself — and that is more or less what happened.

Why this is a feature, not a leak

The tempting read is that the Vikings had one glaring hole and got lucky nobody exploited it enough to matter. I'd argue the opposite: the deep-ball number is expensive because of exactly the same design choices that make the rest of the defense elite.

Flores is sending pressure. Sack rate is up ~40% year over year. QB hit rate is up ~18%. The consequence, mechanically, is that the back end is playing more man coverage and rotating into more single-high looks with fewer bodies deep. That's the trade: rush four-plus-one, live with the occasional busted coverage on the shot play, and in exchange strangle everything underneath. Short throws (+0.017 EPA/play, 5th) and intermediate throws (-0.012, 1st) both got noticeably tougher. Behind-LOS check-downs stayed suffocating (-0.360, 4th). The middle three-quarters of the field got harder to move the ball on.

Opponents responded rationally: they stopped throwing deep. 39 attempts. Fewest in the league. And because those 39 attempts are the ones where the coverage broke — the pressure got home a beat late, or a safety guessed wrong, or a corner turned the wrong shoulder — the plays that did come out of the quarterback's hand connected at a 48.7% clip.

If you're an opposing coordinator, you can look at 6 TDs on 19 completions and think you found the seam. But the volume tells the truer story: quarterbacks in 2025 faced 44 fewer deep looks against Minnesota than they had a year earlier, and the ones they took didn't come with the same turnover risk on the back end. This suggests the Vikings weren't beaten deep so much as they priced deep shots out of the market — and paid, on the margin, for the ones that got through.

Whether that's a sustainable equilibrium in 2026 is a different question. It depends on whether Flores can keep manufacturing pressure at a 9.9% sack rate, and whether the deep-ball INT count reverts from 1 back toward the 10 the same scheme produced the year before. But as a description of what actually happened in 2025 — a top-2 pass defense whose worst tier is dead last in football, by design — this is the shape of the unit, and it's more interesting than the ranking suggests.