The model turns power ratings into fair value in three moves: it converts each team's rating edge into a single-game win probability, simulates the full schedule and playoff bracket thousands of times to produce season-long title equity, then strips the venue margin and blends venues into one consensus number. On the current national title board that process puts Ohio State on top at 11% fair value, with a best price of 13c on Kalshi.
Everything downstream is a translation layer. A power rating is the raw input; fair value is the output a trader can compare directly against a contract price quoted in cents.
What a power rating actually measures
A power rating expresses a team's strength on a single scale, calibrated so the difference between two ratings estimates the expected margin on a neutral field. It is not a ranking or a poll; it is a number designed to be subtracted.
That subtraction is the whole point. Rankings tell an ordinal story, but they cannot be plugged into a probability curve. A rating gap can, which is why the model starts here rather than with human ballots or record-based tiers.
How does a rating gap become a win probability?
The model takes the rating difference between two teams, adjusts for site (home, away or neutral), and passes the result through a curve that maps expected margin onto a win probability. A small edge yields a coin-flip lean; a large edge compresses toward near-certainty.
This is where a slate of matchups becomes a slate of probabilities. Each game on a team's schedule gets its own number, and those numbers are the raw material for everything that follows. Nothing here is a result; it is an expectation about a game that has not been played.
From single games to a season-long title number
Per-game probabilities do not add up to a title on their own. The model simulates the full season many times over, playing out each schedule, seeding a playoff bracket, and running the bracket to a champion. Across those runs, the share of simulations a team wins is its raw title equity.
That simulation captures the parts a single game probability misses: strength of schedule, the odds of reaching the bracket at all, and the compounding difficulty of winning three or four elimination games in a row. It is why a team can grade well week to week yet still carry modest title equity.
Why de-vig and cross-venue consensus decide the final number
Raw market prices are not probabilities. Across a title market the quoted contracts sum to more than 100%, because each venue builds in a margin. The model de-vigs, removing that margin so the numbers read as clean probabilities, then blends venues (here, Polymarket and Kalshi) into a single consensus fair value.
The gap between fair value and best price is the practical read. Ohio State's 11% fair value against a 13c best price shows the markup a trader pays at the current quote. Georgia sits at 7.6% fair value with an 8c best price on Kalshi, a tighter spread near the middle of the board.
Promo codes such as Kalshi's FADE (trade $25, get up to $500) change the cost of entry, not the fair value itself; the model's number is venue-agnostic by design.
Where the model sits on the 2026 title board
The output describes a top tier that is unusually flat. Ohio State (11%), Notre Dame (10.6%), Oregon (10.3%) and LSU (10.2%) sit within a point of one another, with Indiana at 9.2% just behind. That compression is the model telling the reader no single team has separated.
Best prices track the same order but carry markup. Ohio State is cheapest on Kalshi at 13c, Notre Dame at 12c, and Oregon and LSU at 11c each, while Indiana's best price of 11c comes from Polymarket. Reading the price against the fair value, rather than in isolation, is where the pipeline earns its keep.
Prices and the model can both be wrong, and this is analysis rather than financial advice. The value of the framework is consistency: every team on the board is scored the same way, from rating gap to simulated title equity to a de-vigged, cross-venue number.
