Raiders 27, Chiefs 30 — Week 4: Lost It in Thirty-Three Seconds

19 to 17, Raiders, three 35 left in the game, and Kansas City facing third and five at its own 40. At that snap Las Vegas had a 67 percent chance of winning. Patrick Mahomes threw deep left for Tyquan Thornton, and Thornton went 55 yards to the Raiders five. A Las Vegas unnecessary-roughness penalty moved the ball to the two, and Kenneth Walker scored on the next snap. The two-point try failed, 23 to 19. On the kickoff, Dareke Young brought it out to the Las Vegas 22 and lost the ball, and the Chiefs recovered there. Walker ran 22 yards for a touchdown on the next play. 13 Kansas City points in 33 seconds of game clock, and the Raiders' chance of winning went from 67 percent to three. Las Vegas answered with a touchdown and a two-point conversion to make it 30 to 27, watched Kansas City recover the onside kick, and its last drive ended in an interception at the Kansas City 36 with 10 seconds left. It was the Raiders' first loss.

The passing game opens on its success rate, because it held the whole game up: 53 percent of 49 pass plays were successful, meaning they gained what the down and distance asked for. Those plays were worth +8.7 expected points — expected points added being what a snap was worth given the down, the distance and the field position, measured against what that situation was already worth. That is exactly the rate this passing game ran at over the first three weeks. Cousins was not sacked, threw for 365 yards, and spread it: Brock Bowers 11 targets, Michael Mayer 10, Tre Tucker and Ashton Jeanty eight each. The biggest throw came early in the fourth quarter, down 17 to 13. Third and 11 at the Kansas City 46, Cousins threw short right to Bowers, and Bowers got to the 25, 21 yards. +3.2 expected points and 20.6 points of win probability — each team's chance of winning at that moment — the biggest gain in win probability on any Raiders play. That drive stalled and became a field goal, 17 to 16.

The charting explains why the passing game moved the ball without running away with it. 13 of Cousins's 49 throws were charted contested, a receiver and a defender arriving at the ball together. That is 27 percent, against eight percent over the first three weeks. 36 of the 49 were catchable, 74 percent, just under his 75 over the first three weeks, and two were dropped. Play-action came on 13 of 49 dropbacks, 27 percent, twelfth of the 30 charted offenses this week and close to the three-week rate of 24. Off the fake he went nine of 13 for a hundred 33 yards; without it, 22 of 36 for 232. He was outside the pocket on 10 of 49 dropbacks, after six in a hundred across the first three games. Kansas City sent five or more pass rushers on 22 percent of dropbacks, below the 30 percent this offense had seen.

The run game opens on yards per designed run, because it carries the line this show has kept since Week 2: count the box on Jeanty's carries, and if it is light and the runs still go nowhere, that is the story of this offense. Kansas City showed six or fewer defenders in the box on 15 of 26 designed runs, 58 percent, ninth-highest of the 30 charted offenses this week, and well above the 39 percent of the first three weeks. Those 26 runs gained 72 yards, 2.8 a run, under the 3.6 of the first three weeks. 16 of the 26 gained three yards or fewer. The rush plays were worth −6.2 expected points; at the three-week rate, 26 would have been about −5.1. The costliest by win probability came with five 46 left, down 17 to 16: third and two at the Kansas City 30, Jeanty through the right guard for no gain. Matt Gay kicked the 48-yarder for the 19-17 lead. The notebook line holds a fourth week: light boxes, and the runs still go nowhere.

The run defense is the unit that lost the game. 25 Kansas City rush plays went for a hundred 90 yards, two touchdowns and +5.8 expected points. On defense a negative number is the good direction, and over the first three weeks this run defense had been at a rate worth about −5 on that many plays. Walker carried 22 times for a hundred 77. The box was not the change: the Raiders showed a light box on 61 percent of the Chiefs' charted designed runs, right at their 62 percent of the first three weeks. The yards were. Kansas City gained 7.9 a designed run, more than double the 3.8 this defense had allowed. The biggest came with 30 seconds left in the third quarter, the Raiders up 13 to 10 and the Chiefs at their own 16: Walker through the right guard, 76 yards, pushed out at the Las Vegas eight. +4.9 expected points and 22 points of win probability for Kansas City. Two snaps later the Chiefs led. The early stands were real: a third-and-two and a fourth-and-two stop near midfield on Kansas City's first possession, and a goal-line stop on second and two at the two before Thornton scored on third down.

The pass defense opens on what it allowed through the air, because the total tells two stories. 32 Kansas City pass plays were worth +2.7 expected points; at the first-three-week rate they would have been about −5. Only 41 percent of those plays were successful, Mahomes completed 15 of 30, and the Raiders sacked him twice, once on second and three at the three to force a field goal. But three completions — 41 yards to Xavier Worthy on third and 10, 42 and then 55 to Thornton — covered a hundred 38 of his 225 yards. Las Vegas sent five or more rushers on 39 percent of Mahomes's dropbacks, ninth-highest of the 30 charted defenses this week and nearly double its three-week rate of 21.

The other turning point came early in the third quarter, up 13 to 10. Gay made a 41-yard field goal, a Kansas City illegal-formation penalty wiped it off, and the Raiders took fourth and two at the 18 instead of the points. Cousins threw short left for Jeanty, incomplete. −3.6 expected points and 9.7 points of win probability. Three points left on the field in a game lost by three.

Now what repeats. Last week's notebook line was the pass defense, which won in New Orleans on four takeaways during its most catchable afternoon of the season. Against Kansas City the throws got harder to catch: 19 of Mahomes's 30 were charted catchable, 63 percent, under the 67 of the first three weeks. One was interception-worthy, a ball a defender should have caught, and nobody caught it. The defense forced no takeaways. The catchable rate came back down; the takeaways did not come with it. What goes in the notebook: contested throws on Cousins's passes, 27 percent after eight, and a run defense that showed its usual box and gave up more than double its usual yards per run.

The stat lines, scored half-PPR. Cousins, 22.6: 31 of 52, three of them spikes, for 365, two touchdowns and the interception. Bowers, 17.6, six catches on 11 targets for 86 and the 20-yard touchdown. Jeanty, 15.6, 15 carries for 58 six catches for 68. Mayer, 12.2 on eight of 10 for 82. Ian Thomas caught both of his targets for 32 and the opening touchdown, 10.2. For Kansas City, Walker scored 30.4 and Thornton 25.6.

For anyone joining: wins over Miami, at the Chargers and at New Orleans, and now a three-point loss at home to the unbeaten Chiefs. Three and one.

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

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