Texans outplayed the Chargers by three touchdowns' worth of snaps
Houston won 44-22, and the snaps say the real gap was 20.7 points, so the score hid nothing in the Chargers' favor.

The verdict
Played like that, the Chargers win 0% of the time. The snaps gave Houston a deserved win probability of more than 99%, and the deserved margin was Texans by 20.7. The final was 44-22, a 22-point gap. This was earned, start to finish.
There is no luck argument to make for Los Angeles. A team that loses by about what the snaps predicted did not get robbed.
What the score hid
The score hid very little, and what it did hide cut against the Chargers. Los Angeles averaged -0.14 EPA per snap. EPA is expected points added: how much each play moves the scoreboard forecast. Houston averaged 0.28, a 0.42 gap per snap.
Success rate told the same story. The Chargers stayed on schedule on 36% of snaps, the Texans on 45%.
Luck points, which measure how much of a result came from bounces rather than play, ran -3.9 for the Chargers and 0.6 for the Texans. Bad luck cost Los Angeles a few points, and it still lost by three touchdowns.
The swing play was Houston's. With 4:37 left in the second quarter, C.J. Stroud's name is not in the brief, so the play stays as written: (Shotgun) 7-C.Stroud pass deep right to 1-T.Dell for 63 yards, TOUCHDOWN. It added 18 points of win probability for Houston, the largest single move of the game.
The calls
The fourth-down log is mostly clean. Three field goals matched the model, and the second-quarter punt from 78 yards out was clear and correct. The one miss came late in the first quarter: fourth-and-7 from the Houston 41, a punt where the model preferred a field-goal attempt. It graded as a lean, worth 0.7 expected points.
The flags hurt more. A face mask on the Chargers in the first quarter, 15 yards at 13:14, swung win probability 8 points toward Houston. A second-quarter false start at 1:24 cost another 4. Houston gave some back, with a face mask for 15 yards (3 points) and a false start (3 points) of its own, but the net ran Houston's way.
Jim Harbaugh's problem is not this game. It is the pattern. His Chargers have faced 42 graded fourth downs, and the model recommended going on 8 of them. They went once. The Chicken Index, the share of go-for-it spots where a team kicks or punts, sits at 88%, second most timid in the league. Those decisions have cost 5.8 expected points, with 2 clear misses. Against Houston the cost was small. Across the season it is not.
The trend
The season record reads 0-8, and the snaps say the team deserved one win. That puts the luck sheet at -1 win, 28th of 32 on the luckiest-first list. Total luck points are -4.9, including -1.6 on fumble recoveries. Some of the losing is bad luck. Most of it is not.
The Chargers are 0-3 in one-score games, so they have not even collected the coin flips. Their net EPA per snap is -0.224, last in the league. They have taken 66 penalties for 540 yards, and the flags have cost them 39% of win probability on net.
The luck should regress toward zero. Fumble recoveries and opponent kicking tend to even out. The play is a separate matter: with the worst net EPA in the league, there is little for the luck to correct.
Next is a road game at the Baltimore Ravens on Monday, Nov 16, at 8:15 PM ET on ESPN.