O'Connell rebuilt the Vikings offense around Kyler Murray, and Detroit is the first real test

Minnesota walks into Detroit at 6-0 with an offense that ranks 29th in efficiency. Kyler Murray did not cause that gap. Kevin O'Connell's play sheet did, and Sunday asks whether it holds against a division rival that has seen all of it on tape.

The passer is not the headline

Murray's season line is solid and unspectacular: 66.4% completions (83 of 125), 812 yards, three touchdowns, two interceptions and an 85.8 passer rating. Compare that with 2024 in Arizona, when he completed 68.8% and finished with a 93.5 rating. Murray is not playing better in Minnesota than he did there. The wins are coming from how the offense is built around what he does.

The offense scores 20.8 points a game, 24th in the league. Teams with that profile usually have a few close losses. Minnesota has none, which tells you the margin is thin and the design has to keep working.

Play-action is the structure

The Vikings use play-action on 32.3% of dropbacks, second in the NFL. That is the clearest sign of what O'Connell has done. Play-action gives a quarterback a defined read, a moving pocket and a defense with its eyes on the run fake. For a passer like Murray, who is at his best when he can set his feet or move off a clean launch point, that structure removes some of the guesswork that drive-killing dropbacks create.

It also fits the run game. Murray has 132 rushing yards on 25 carries, a little over five yards a carry. Those are not designed quarterback runs for show. They keep a defense honest on the edge, which is what makes the fake worth believing. O'Connell has put the threat of Murray's legs inside the same plan as the play-action calls, and each one feeds the other.

What Detroit changes

The structure works best when the offense stays ahead of the defense. Detroit knows the tendencies. A team that uses play-action at this rate gives an opponent a clear target: stay disciplined against the fake, keep a defender on Murray's escape lane and make the quarterback win from a standard dropback. Minnesota's 29th-ranked efficiency says it has not yet shown how it would answer that.

That is the test for O'Connell. The play-calling has carried an offense that does not yet look like a top-tier unit, and the schedule has not asked for more. A Lions defense that does not bite on the first fake forces him to call something else. If Murray has to throw into coverage on a normal down, the numbers suggest that is where the 6-0 start gets harder to protect.

Whatever the result, the lesson from the first six weeks stands. Minnesota did not simply add a quarterback. The coach changed the offense to fit him, and the record reflects the design more than the player.