UAB at Louisiana. Lean: UAB +7.5. Lean: Under 56.5.

Spread.I make Louisiana roughly -6.5, versus Edge Labs’ -14 and the market’s -7.5. The model’s raw margin is only 12.1 before a calibration that stretches favorites, and its own testing shows no ATS edge. UAB’s preferred style supports the underdog: 100 rushes against 47 passes, 36:28 average possession, and 460 rushing yards through two games. That can shorten possessions and make the hook above seven valuable. Louisiana is still the more explosive team, Lunch Winfield is completing 72.7% and adds a major rushing dimension, so this is only a modest position. The move from approximately Louisiana -10 to -7.5 has already incorporated UAB’s competitive start and Louisiana’s receiver problems; FOX’s displayed split also showed larger money on UAB, though its underlying sample is not disclosed. Shelton Sampson’s status for Saturday remains unverified after missing USC, while UAB’s Jaylen Elder is questionable but has no 2026 production. Take UAB +7.5, preferably only at 7.5 or better.

Total.My total is about 54.5, below both the 56.5 market and Edge Labs’ 59.9. UAB’s 100-to-47 run/pass balance, 36:28 possession average and only two touchdowns on six red-zone trips create an under-friendly combination of clock consumption and field goals. Louisiana’s missing or uncertain receivers further lower its passing ceiling. The market has already made the obvious adjustment, DraftKings fell from 59.5 to 56.5, so the remaining edge is modest. The serious over case is Louisiana’s 6.9 yards per play, six touchdowns on seven red-zone trips, Winfield’s dual-threat ability and nonrestrictive weather. Still, UAB’s preferred game script should limit possessions enough to favor under 56.5.

Where our number looks wrong.The -14 projection is too bullish on Louisiana. It converts a 12.1-point raw margin into 14 despite an unproven calibration, does not recognize UAB’s clock-shortening run profile, and cannot price Louisiana’s depleted receiving group. Its early ratings also blur opponent context: Louisiana’s data are one FCS game and one game against USC. The totals model’s 59.9 similarly looks high because it misses UAB’s pace, red-zone field-goal tendency and the receiver uncertainty.

Spread.Our model prices Louisiana -14, but that number carries no proven ATS edge and the calibration map stretches a modest rating gap (roughly 9.6 pts plus 2.5 HFA) into a double-digit favorite. That is ~2 pts of reverse line movement toward the dog with no reported Louisiana boost to justify it, which usually means sharper money on the underdog. Louisiana is the better team at home with a healthy dual-threat QB (Winfield) and RB Darrell Smith back, but three receivers show as banged up on the Covers scan, capping the passing game against a UAB team that just beat a comparable Sun Belt side (ULM) by six. Taking the points with the market rather than fading it: UAB +7.5.

Total.The market has drifted down, not up: DraftKings 59.5 to 56.5, MGM/FanDuel 57.5 to 56.5, all landing at 56.5. Weather is a non-factor (clear, ~89F, ~6 mph wind at Cajun Field), so this is a personnel/pace read. Louisiana's three dinged receivers cap the explosive passing that drove 38 on Lamar, and UAB is a ground-first offense (Rod Robinson II carrying the load) that shortens games. Adding implied scores (~30 Louisiana, ~23 UAB) lands near 53, under the number. Lean under 56.5, but only a modest one given both defenses have already surrendered 40-plus this season.

Where our number looks wrong.The model's -14 is far outside a market that has moved the other way to -7.5; the raw rating gap plus HFA is ~12, and the calibration stretch on top of that overstates a mid-tier Sun Belt home favorite. The rating also cannot see Louisiana's thinned WR room or that UAB just handled a peer Sun Belt team, both of which argue the true number is closer to the market than to -14.

Spread.I'm overriding that with a lean toward UAB. Three Louisiana wide receivers (Strother, Chaney, Sampson Jr.) have been out since Wed 9/9 with shoulder/shoulder/knee injuries (Covers.com, 9/16), thinning the exact unit that has to replace production against a defense that just gave up 49 to USC's own leaky group. Louisiana's 1-1 record is inflated by a 38-7 walkover of FCS Lamar; its one FBS look (at USC) was a 19-point loss with defensive breakdowns. UAB is 1-1 with a competitive road loss at Illinois and a week-2 win over ULM. Taking the number the whole market is buying.

Total.That lines up with personnel: Louisiana is missing three wide receivers (Strother, Chaney, Sampson Jr., all out since 9/9), which should crimp the passing offense that scored 30 on USC. Weather is a non-factor, clear, 89F, 6 mph wind. Both defenses have been generous (Louisiana allowing 28 ppg, UAB 31 ppg), arguing some for over, but that's already priced into a still-high 56.5, so I'll side with the market drift and the receiver attrition.

Where our number looks wrong.The model's -14 is likely overstating Louisiana here for two stacked reasons the dossier itself flags: a flat 2.5-point home edge at every venue, and a calibration map that stretches large raw margins even bigger. Neither term sees personnel, and the model has zero read on Louisiana's WR room being down three bodies since 9/9, exactly the kind of skill-position hit that shrinks a home favorite's margin. The model also can't see that the market has spent four straight days buying UAB back from double-digit dog to a touchdown, which is real information the rating never ingests.