Here is our Gardner-Webb vs Marshall prediction for Saturday, September 26, 2026: Marshall wins at home, and the model's projected margin is Marshall -24.6. The moneyline on Marshall sits at -3500, and the tier on this game is PASS. That is an official-card PASS, which means we are not backing it and there is no official action from us.
Picking the winner is the easy half. Whether the bet is worth making is a separate question, and today we are not making it.
Why the model is on Marshall
The model's read starts with how Marshall scores. Its win path is sustained offense at home behind Carlos Del Rio-Wilson, who is the last starter at quarterback and a readable one. The Thundering Herd have beaten FBS opposition 28-26 and 30-24, and the desk weighs those results more heavily than the 0-45 loss at Penn State.
Those are narrow wins, and the model says so. They do not establish trench dominance. Still, a team that can win two games against FBS opponents is a different proposition from a Gardner-Webb side that is stepping up in class, and the projected margin of 24.6 points reflects that gap.
The case against Marshall
The concerns are real, even if they do not flip the pick. Marshall used a three-quarterback rotation earlier in the year, and a dated shoulder and collarbone availability report leaves questions about durability and offensive continuity. If the quarterback situation wobbles, a big spread gets uncomfortable quickly.
Gardner-Webb has its own path. Cole Pennington is the engine: the September 19 official report credits him with 309 passing yards and two touchdowns. The Runnin' Bulldogs also lost 23-24 at FBS Liberty, a one-point result that says this team can hang around against a higher level of opponent. Bennett is in an interim role on the Gardner-Webb side, and we are not going to invent a coaching bounce or a scheme change from that. We simply treat it as one more layer of uncertainty.
The number that matters: -3500
A -3500 moneyline means a bettor risks $3,500 to win $100. Even if Marshall wins, as the model expects, the payout is tiny compared with the money at stake. A single upset, or a close finish that turns into a coin flip late, wipes out a long run of small wins.
The bot's own risk notes describe the scenario that worries us: Pennington reproduces his documented passing production, Gardner-Webb repeats the competitiveness it showed at Liberty, and Marshall's quarterback disruption returns. That combination turns the game into another close finish. It is not the model's central case, but it is a live one, and it is the reason a 24.6-point projection and a bet we want to make are not the same thing.
Where we land
The model's baseline carries no adjustment for the quarterback or coaching questions, because those overlays call for caution but not a defensible number. We also make no claims about weather, rest or scheme, since we have nothing solid on any of them.
The tier is PASS. We think Marshall is the more likely winner, and we are not backing this game.
The verdict
Verdict: The model projects Marshall -24.6 and has the Herd winning straight up. The moneyline is -3500, the tier is PASS, and this is not a bet we are making. Read it as a call on who wins, not a recommendation to lay the price.
- Event · College Football Week 4 · Saturday, September 26, 2026
- Market · Marshall -3500 — 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.