The longshots with a real model edge are the 5c-and-under contracts, where the best Polymarket price sits at or below the model's fair value: Georgia, LSU, Texas A&M, Alabama and Oklahoma. Each of those teams prices cheaper than the model's de-vigged estimate, which is the opposite of what happens at the top of the board.
That distinction matters. A longshot is only interesting when the price undercuts the probability, not when it simply looks small. On the current national title board, that condition holds across the entire bottom tier.
Which longshot CFB contracts show a model edge?
The edge lives where best price is at or under fair value. Georgia's fair value is 5.6% against a 5c price. LSU's fair value is 5.4% against 5c. Texas A&M sits at 3.1% fair value against 3c. Alabama prices 2c against a 2.5% fair value, and Oklahoma prices 2c against 2.2%.
Every one of those contracts trades on Polymarket, which holds the best price across the board in this range. The consistency is the signal: this is not one stray quote but a full tier priced a fraction under the model.
Contrast that with the favorites. Notre Dame's best price is 12c against a 11.8% fair value, Ohio State is 11c against 11.1%, and Oregon is 11c against 10.9%. The top of the market prices slightly rich, while the tail prices slightly cheap.
How big is the gap on Georgia and LSU?
Georgia and LSU headline the tier because they carry the highest fair value among the discounted names. Georgia at 5.6% and LSU at 5.4% both round to a 5c price, so the contract cost lags the model's probability rather than leading it.
The absolute gap is small, measured in tenths of a percentage point, but the direction is what the model tracks. A buyer paying 5c for something the model values at 5.4% to 5.6% is on the correct side of the vig, unlike a buyer paying 12c for a 11.8% estimate.
Why the 2-cent tier still carries an edge
Alabama and Oklahoma sit at the floor of the tradeable board, both priced 2c on Polymarket. Alabama's fair value is 2.5% and Oklahoma's is 2.2%, so each contract prices below the model's number despite the long odds.
At this level the rounding works in the buyer's favor: a price that cannot tick below 2c will sit under any fair value between 2% and 3%. That is a structural feature of a discrete-cent market, and it is where longshot edges tend to hide.
The best-price picture confirms how compressed the tail is. Five teams cluster within a 3c range, which is why small differences in fair value drive the whole read.
How to read these longshot edges
A model edge on a 2c to 5c contract is not a verdict that any of these teams will win the title. The base rate remains low, and a fractional gap between price and fair value can vanish on a single news cycle or liquidity shift.
What the numbers describe is relative value: within the longshot pool, the market is asking less than the model's probability, while the favorites ask slightly more. Both the price and the model can be wrong, and none of this is financial advice.
For traders comparing venues, Polymarket is the reference price across this tier, with the code TGSWC available at signup. The takeaway is narrow and data-led: the cheapest contracts on the board are also the ones priced under fair value.
