Colts, Week 2 — What Repeats: Play-Action Went 25th to 2nd

Last week the Colts faked a handoff less often than most teams. On Sunday night they faked on 16 of 33 dropbacks. That is 48 percent, the second-highest play-action rate of the 32 teams this week, behind only San Francisco, with the league median at 24 and a half. Against Baltimore it had been six of 34, 18 percent, 20-fifth in the league. Across all of last season it was 28 percent. So in one week this offense went from well under its own norm to far over it, and the fake did real work. Daniel Jones threw 15 play-action passes and completed 12 for a hundred 25 yards. His 16 other throws went 10 for 85. More than half of his 210 passing yards came off play-action, after four play-action throws for 15 yards against the Ravens. The Recap had the passing game swinging from deep in the red to +6 expected points. The charting says where a large share of the yardage lived.

Now the figure that did repeat, and it repeated exactly. Nine of Jones's 31 throws were contested, meaning a receiver and a defender arrived together. 29 percent, the third-highest rate of the 32 teams this week, behind Jacksonville and Green Bay. Against Baltimore it was also nine of 31, the highest in the league that week. Across last season Jones's contested rate was 16 percent. And it does not look like the opponent. Kansas City's defense saw only eight percent of throws contested across all of last season. Two defenses, the same share of throws into traffic, close to double his own norm. It is one of several figures on this offense that sat on the same side of last season in both games, and the widest gap of them. The rest of his ball charting was steady. 23 of the 31 were catchable, 74 percent, tied for eleventh with Buffalo, Cincinnati and Denver, a little above the 72 he posted last season. One was charted interception-worthy, a ball a defender should have caught, three percent of throws, tied for eleventh with Baltimore and the Rams, and the same one-in-31 as last week. The game had one interception, and it was that throw. And no drops, after three last week. Indianapolis was one of eight offenses with none.

The run game went the other way from the passing game. Kansas City showed a light box, six or fewer defenders, on 15 of the Colts' 30 designed runs. Half of them, twelfth-highest of the 32 this week. Last week Baltimore did it on 63 percent, and last season this offense saw a light box on 62 percent of its designed runs. So fewer light boxes than usual, though the average box barely moved, six and a half against 6.4, and the 30 designed runs went for a hundred 12 yards. 3.7 a run, nineteenth in the league this week, after 5.2 and fifth against Baltimore, and after about four and a half across last season. One week the designed runs were near the top of the league. This week they sat in the bottom half.

The rest of the scheme mostly held its shape. Screens on three of 33 dropbacks, nine percent, twelfth in the league, after nine percent last week and five across last season, so that one is up two weeks running. Motion on 42 of 85 charted plays, 49 percent, seventeenth and just under the median, after 54 last week and 43 last season. No-huddle went to zero. Indianapolis was one of six offenses that never used it this week, after 13 percent of snaps last week, when it trailed for most of the afternoon. Kansas City sent five or more pass rushers on 11 of 33 dropbacks, 33 percent, sixteenth, close to the 30 percent this offense saw last season, and Jones was not sacked.

The defense threw out last week's rush plan. It sent five or more rushers on nine of Kansas City's 51 dropbacks, 18 percent. Of the 32 defenses this week, only Cincinnati sent five or more less often. Against Baltimore it was 38 percent, seventh-highest. Across last season it was 28. Against that, Patrick Mahomes put 35 of 47 throws in a catchable spot, 74 percent, the tenth-highest allowed this week, against 71 percent for everyone who threw on this defense a year ago. Five of his 47 were contested, 11 percent, nineteenth, down from 20 percent against Baltimore and 12 across last season. One interception-worthy ball, one drop. Last week the defense sent more, and the windows were tighter. This week it sent less, and they were wider. Two opponents, two plans, and neither has earned the word yet.

What did repeat on this side is how offenses choose to attack it. Kansas City used play-action on 19 of 51 dropbacks, 37 percent, fourth-highest in the league this week. Baltimore did it on 34 percent last week, fifth-highest. Offenses facing this defense last season used it on 22 percent. Mahomes went 15 of 18 for a hundred 25 yards on the fake. And the designed runs repeated too, in a way that matters more. The box changed completely. Last week the Colts showed a light box on 29 percent of designed runs, well under last season's 57. This week it was 14 of 27, 52 percent, eleventh-highest, right back near last season. Kansas City's 27 designed runs went for a hundred 35 yards, five a run, the fourth-highest allowed this week. Baltimore's went for 5.4, third-highest. Last season this defense allowed 3.8 per designed run. Heavy boxes one week, normal boxes the next, and five yards a run both times.

The stat lines have settled, so here they are, scored half-PPR. Jonathan Taylor led the Colts with 27.2, 24 carries for 92 two touchdowns, +4 catches on four targets for 40. Jones, 11.4: 22 of 31 for 210, a touchdown, an interception, and 10 yards on three carries. Tyler Warren caught six of seven targets for 21 yards and the touchdown, 11.4. Josh Downs, seven catches on nine targets for 72, 10.7. Laquon Treadwell caught all three of his targets for 61. For Kansas City, Mahomes had 29, 32 of 47 for 382 three touchdowns. Kenneth Walker ran 24 times for a hundred 17 and caught six of 10 targets for 61, 20.8. Travis Kelce, nine catches on 11 targets for a hundred one and a score, 20.6.

A word on the measuring stick. Everything above is set against last season and against last week, because through Week 3 that is all there is. Last season was a different roster, a year younger, and in places a different plan. Last week is one game. Those comparisons are good at flagging what is unusual about Sunday. They cannot yet say what is true about this team, and two games do not change that. From about Week 4 this show starts measuring the Colts against their own run of weeks.

For anyone new, here is the thread. A year ago this was an eight-and-nine team built on its run game, third in the league by expected points per play, with a pass defense that ranked nineteenth. This season opened with a 41 to 23 loss to Baltimore and now a 33 to 30 overtime loss in Kansas City. The notebook from last week had three lines, and all three reversed at least in part. The drops vanished. Play-action went from missing to second in the league. The heavier rush plan was dropped for one of the lightest in the league. What goes in now is what held. Jones's throws were contested 29 percent of the time in both games, close to twice his norm. Opponents have faked against this defense at a top-five rate both weeks. And designed runs have gone for five yards a carry or more against it twice, with two very different boxes. So on Sunday, watch the play-action against the Colts' defense and what the run does after it. That is the pattern two games have drawn.

That's your Colts — every game, by the numbers. This is Muffed: we do the same deep dive for all 32 teams. Follow the Colts show wherever you're listening, so next week's recap lands the moment it's out. No theater. Just the football.

And if you play fantasy, there's a Muffed show built around your exact roster — free at muffed.ai.

_Data via nflverse. Charting: FTN Data via nflverse._