Los Angeles Chargers vs Buffalo Bills prediction: 5 of 5 AI models back Buffalo Bills

The shape of the game

All five experts land on the same side, and the reasoning converges on a coherent picture rather than a coin-flip consensus: this is a matchup of trajectories, and both trajectories point decisively toward Buffalo. The Bills enter 2-0 having scored 36 and 41 points behind an offense that has been the most efficient early-down unit in the league through two weeks. The Chargers arrive 0-2 having scored exactly 14 points in consecutive home losses to Las Vegas and Arizona — not against elite defenses. Two 14-point outputs against middling opposition is the single most damning data point on the board, and it is the number every model independently flagged as decisive.

Where the real separation lies

The panel's sharpest shared insight is that the Chargers' offensive line and skill-position attrition is the true driver, not the quarterback comparison. Los Angeles is missing both starting-caliber interior linemen (Pipkins, Awosika) on top of Biadasz already being unavailable, and the tight end room is down to a third option with Njoku on IR and Kolar out. Herbert is being pressured into the exact execution failures — turnovers, failed red-zone trips, three-and-outs — that produced those 14-point afternoons. Now that patchwork front travels three time zones east on six days' rest to face a Bills pass rush that has already notched seven sacks, while Buffalo sits on nine days of rest after a Thursday win. Rest, travel, injury, venue, and form all point the same direction, which is why the panel is willing to price this near a true mismatch.

The honest counter-case

Where the models properly hedged is on Buffalo's defense, which has allowed 31 points in both games. This is an offense covering for a leaky secondary, not a complete team, and Herbert with McConkey and Gadsden has enough talent to exploit the Bills' boom-or-bust coverage if the protection holds even intermittently. A Harbaugh-coached team is also structurally built to shorten games and keep scores respectable, which caps the likely margin rather than the winner. That is why the panel caps Buffalo's win probability in the mid-70s rather than treating this as a two-touchdown certainty, and why the most likely scorelines cluster in the comfortable-but-not-blowout range.