The Vikings Fell to the Boom of the Lions Run Game

At halftime in Detroit, it was going pretty well for the Minnesota Vikings’ defense. They had allowed just six points against the dynamic Lions’ offense and were dominating against both the run and pass. They had sacked quarterback Jared Goff six times, and the Lions had run the ball 16 times for just 66 yards.

Jahmyr Gibbs, in particular, was bottled up, with just 18 yards on his first 10 carries, before a 19-yard run with 12 seconds left in the half, when the Vikings were mostly just trying to prevent a field goal. At the end of the second quarter, the Vikings were allowing -0.380 EPA/rush and just a 25.0% success rate on rushing plays.

Prior to this game, the Vikings were among the league leaders in both EPA and success rate against the run, at -0.157 EPA per carry and 35.1%, ranking fourth and first, respectively. They had allowed just 3 first downs against the run.

Even late in the 3rd quarter, the Vikings’ defense held up against the run. Prior to the Lions’ last play of the quarter, the Vikings had allowed 91 yards on 22 carries, excluding a lateral to WR Amon-Ra St. Brown. When removing the effect of dead-ball penalties, they had allowed only seven successful runs, five first downs and -0.283 EPA per carry (per nflverse EPA).

Then, disaster struck. Gibbs broke off a 30-yard touchdown run, and the defense never recovered. In the 4th quarter plus that one play, the Vikings allowed 131 yards on 14 carries (excluding two Jared Goff kneel-downs) and eight total first downs. 

To Vaki and Gibbs alone, they allowed 127 yards on 12 carries (10.6 per carry), with nine successful runs. That more than canceled out the Vikings’ first-half performance, as they allowed 0.539 EPA per run and a 75% success rate in the second half.

That unraveling, along with continued offensive ineptitude, cost the Vikings the game, and they lost to the Lions 19-15.

What happened?

Some of the Vikings' run issues can be chalked up to Jahmyr Gibbs being awesome, but not all of them. The truth is that the Vikings’ defense, despite excellent numbers, had shown some cracks against the run in prior weeks. 

Not just in the Colts game, where they allowed a middling-to-below-average 0.028 EPA per carry and a 43.8% success rate on the ground, but in allowing 14 runs of 10+ yards over the first 7 weeks of the season. 

On 150 carries, that 9.3% rate of explosive runs allowed was still above average — 13th in the NFL, but not elite. The three other best run defenses in the league were all under 7%. After the Lions’ game, that number is now 12.6%, 27th in the NFL.

It’s noteworthy that it took the Lions — who are currently 13th in rush EPA/play at -0.083 but an excellent 3rd in success rate at 47.0% — to finally gash the Vikings’ defense. 

It’s difficult to stop Jahmyr Gibbs, but the Vikings have to play him again and also have Bijan Robinson on their schedule. The Lions’ offense is awesome, but so are the Buffalo Bills and San Francisco 49ers, who are two of their next three opponents. The Bills (2nd) and 49ers (1st) are the only teams with a higher rushing success rate than the Lions, and also have a significantly higher rush EPA per carry than the Lions do (3rd for the Bills, 4th for the 49ers). This is a very difficult stretch of schedule for the Vikings.

A lot of ink has been spilled about the Vikings’ defense this year, and rightfully so. Even with the second-half run-defense collapse, they’re still probably the best defense in the league. They rank first in points per drive, opponent-adjusted points per drive, EPA per play allowed, opponent-adjusted EPA per play, and rank third in defensive DVOA.

They’re unique not just in blitz rate, but also in the personnel they put on the field. The Vikings run “base” personnel (defined as 4 or fewer DBs on the field) on 64.2% of plays, which is more than double the league rate of 30.5%, per SumerSports. The Dolphins, the 2nd-place team, run the ball at a comparatively low 51.7%.

The Vikings rank first in the rate of putting 8+ players in the box, implying that they “stack the box” at a high rate. However, that doesn’t mean that they are actually ahead in what I’ll call “box count,” the number of defenders in the box vs. the number of blockers on a potential run play — this data seems limited to what telemetry on offensive position can provide, which is locked behind some pretty significant paywalls.

When a defense designs run fits, they have to account for every gap the offense has. Gaps are the spaces between the blockers, plus the outside edges of the line, so there is always one more gap for an RB to run through than the number of blockers. 

Therefore, if a defense is +1 in box count, they have a defender for every gap. If the box count is even, the defense has to ask a player to two-gap or multiple players to gap-and-a-half as a way to close that uncovered gap.

The best proxy for box count is probably safety positioning. Generally, defenses need to cover each player lined up out wide with a defender of their own. In 12 personnel with two receivers out wide, you would have two CBs covering them. 

That would leave 7 blockers — 5 offensive linemen and 2 TEs. In the Vikings’ base defense, they would default to having 3 IDL, 2 Edges, and 2 LBs in the box, so 7 players. That would mean the box count is even. You can’t pull a CB from out wide to increase your box count, so the advantage would have to come from a safety coming down into the box. 

Against 3 WR sets, you theoretically could have the slot receiver either covered by someone in the box (the Vikings sometimes do this by dropping a player off the line of scrimmage in their all-up blitz looks), or have the player covering the slot defender factor in the run defense (as an “apex” player) rather than a safety, but still the majority of the time, the location of the safeties tells you how many players are in the box and thus able to take part in the run fit.

The Vikings, who play Cover 2 and Quarters a lot and are obviously constantly rotating, play their safeties relatively shallow. Still, they play a lot of 2-high. They show 2-high 63.0% of the time, 14th in the NFL and slightly above the 61.3% average. 

Interestingly, that actually increases against run plays (68.0%, 8th in the NFL) and decreases against the pass (59.1%, 20th in the NFL). They consistently drop back to 2-high even if they show blitz. Post-snap coverage data is only available for passing plays (for obvious reasons), but the Vikings run 2-high 66.7% of the time post-snap, which is 4th in the NFL and well above the league average of 50.8%.

That leads to an interesting situation. Because the Vikings run 2-high more than they present it, they frequently rotate back into coverage, which means (at least conventionally) it’s more likely the safeties step backward at the snap.

That said, because the Vikings run so many funky coverages, like inverted coverages and non-traditional Tampa-2, many “two-high” looks will have a safety stepping forward. Nevertheless, safeties in run defense are likely driving on the run from depth rather than from inside the box.

Despite the box numbers above, the Vikings do not consistently “gap out” opposing defenses by having a defender in the fit for every gap, unless you count the deep safeties.

This is where I think the variance in the Vikings’ run defense comes in. They often rely on a player to 2-gap, or at least play “gap and a half”, while also depending on safeties to come from depth. They also stunt heavily in run defense. 

Those tendencies both create potential problems; we’ve seen interior defenders struggle to hold up against double teams and even single blocks in run defense, it’s difficult for safeties to tackle from depth (and it costs yards even if they make the plays), and stunts can leave wide open paths if they play into the run the offense has called.

None of those downsides mean the Vikings should stop playing this way; they’ve obviously been very successful despite those risks. Playing so much base means that they have bigger bodies than most other teams, helping them hold up against the run. 

Their defensive line is quick to create disruption; linebackers Blake Cashman and Eric Wilson are extremely violent at LB, and the safeties are sound coming up in run support. Generally, the defense works well against the run. It just happens to have a boom-bust element to it. 

The Lions got the bust (or boom, I suppose, from the defensive perspective) side of that in the first half against the Vikings, but they hit the boom in the second half. Let’s go to the tape to break down how it happened.

Throughout the good and bad of the Vikings’ run defense on Sunday, I tried to categorize why each play went the way it did. I came up with 4 overarching categories. Plays were primarily driven by Flores’ unique defensive strategy, an individual defender making a good play or mistake, the opposing RBs making plays, and some really unique offensive designs by the Lions.

That’s roughly sorted from good to bad, and because that’s how the team’s run defense worked throughout the game, that’s how we’ll go through them.