Every Louisiana vs Southern Miss prediction this week starts with the same split screen. The market has priced Louisiana as a 10-point favorite on Saturday, October 24, 2026, with a moneyline of -385 attached to the Ragin' Cajuns. The model, working off its own numbers, also likes Louisiana straight up, but by a smaller margin than the spread demands.
That gap is the whole story. This is officially a PASS on the card: no action, no bet going up for this one, regardless of how the game actually plays out.
Why the model still likes Louisiana
Strip away the spread and the straight-up pick is straightforward: the model has Louisiana winning by 7.9 points. That lean comes largely from quarterback stability. Louisiana's Walter Winfield has seven starts under his belt, while Southern Miss turns to Hunter Lyddy, who has four. That experience gap makes Southern Miss's offense more readable on paper, especially with starter Trey Hampton carrying a verified Out designation as of October 22. Limited throwing practice during the week doesn't change that; it neither clears Hampton to play nor hands Southern Miss a proven backup plan at the position.
Louisiana's path to covering, in the model's view, runs through sustained offensive production against a Southern Miss defense that gave up 55 points to Troy this season. Louisiana's own 33-6 result over Troy gets cited as a supporting data point, though it comes with real caveats: different venues, and a pair of lopsided scores against the same overmatched opponent that don't by themselves prove anything about the trenches.
The case for Southern Miss covering anyway
This is where the PASS earns its keep. Southern Miss's more relevant result isn't the Troy game, it's a 21-24 road loss at Tulane, a performance that was competitive rather than a blowout. If Southern Miss plays to that level again at home, a 10-point line gets uncomfortable fast, even if Louisiana wins outright.
The bigger lever is Louisiana's defense, which is dealing with season-long absences for McPherson, Darbonne, Holts and Spencer. That's a depleted rotation for a unit that's being asked to close out a double-digit favorite, and it's the core of the risk case here: Southern Miss repeating its Tulane-level defensive performance at home while Louisiana's defensive absences and uncertain receiving depth turn what should be a modest offensive advantage into a close-game upset. Any unconfirmed reports of Louisiana receivers returning to the lineup don't change that calculus until they're established on the field.
Southern Miss isn't whole either. The team is without Jahmir Davis, with Rogers and Nunnery both listed Out and Briggs lost for the season, which squeezes both protection and backfield depth. None of that, on its own, is being used to add an extra penalty to Southern Miss's number. It's already baked into results like the Tulane game, where those absences were already a factor.
The number that matters
Louisiana -7.9 against a market line of Louisiana -10 is roughly a two-point gap, and that's before accounting for a defense playing without four regulars and a receiving corps the model isn't crediting with a clean recovery. A model number sitting under the market's spread on a double-digit favorite is exactly the kind of setup that keeps a side off the card even when the straight-up outcome looks likely to go the model's way.
None of this is a case against Louisiana winning the game. It's a case against laying 10 points to do it.
The verdict
This is a PASS, not a bet. The model's own number sits under the market's spread, and Louisiana's defensive absences add enough uncertainty that the margin isn't one we're backing at -10. We're not making this bet, straight ahead, and nothing further is implied about why.
- Event · College Football Week 8 · Saturday, October 24, 2026
- Market · Louisiana -10 -385 — recorded when the pick was posted
- Model · Heismann, Botcappers' AI College Football model — how it is built and what it does not know
- Record · every pick is timestamped before the event and graded in public, win or lose
- Statistics · the model's own match/game database, current to the posting time above. Figures a reader cannot open here are internal to that database, not published third-party data.

