Redshirt
Analysis

Why Prediction-Market Prices Move: Reading the Signal

Prediction-market prices move on three forces: liquidity, news and model fair value. Here is how to read which one is driving a CFB futures price.

By The Model Desk · 2026-09-19
Analysis
vsvsvs
Key takeaways
  • Prediction-market prices move on three forces: order-book liquidity, incoming news, and drift back toward the model's fair value.
  • Texas prices at 14c on Polymarket against a 12.2% fair value, the clearest case of price sitting above the model on the title board.
  • Georgia, Ohio State and Notre Dame all trade near 12c against fair values of 11.6%, 10.7% and 10.7%, prices tracking the model closely.
  • Ole Miss trades 5c against a 3.3% fair value, a tail contract where thin liquidity widens the price-to-model gap.
  • Kalshi carries the sharpest prices on title favorites while Polymarket tends to hold the cheaper tail contracts.

Prediction-market prices move on three forces, and reading a chart well means naming which one is in control: liquidity, news, or the pull back toward the model's fair value. Liquidity sets how far a single order can push a price; news resets what the price should be; and fair value is the anchor that says whether a move is signal or drift. When those three agree, price sits on top of the model. When they diverge, the gap is the trade.

What actually moves a prediction-market price?

Every contract price is a probability in disguise. A 12c contract implies roughly a 12% chance the market resolves yes, before fees and spread. That number changes for only a few reasons, and separating them is the whole job.

Liquidity is the mechanical force. In a thin book, a modest order clears several price levels and the last print jumps, even though nothing about the team changed. News is the fundamental force: a result, an injury or a roster move that genuinely shifts the odds and should move the price. Fair value is the reference: the model's independent estimate of the true probability, which lets a trader judge whether a given move was earned or just liquidity noise.

The model's title board gives a clean set of reference points. Texas anchors the top at 12.2% fair value, with Georgia at 11.6%, and Ohio State and Notre Dame level at 10.7%. Those are the numbers a live price is measured against.

How liquidity shows up in the spread

Liquidity is visible in where the best price sits relative to fair value. On the favorites, the two track tightly. Georgia prices 12c on Kalshi against 11.6% fair value; Ohio State and Notre Dame both price 12c against 10.7%. Deep books absorb size without dislocating, so price and model stay close.

The tail behaves differently. Ole Miss trades 5c on Kalshi against a 3.3% fair value, a wider relative gap because longshot contracts carry thinner order books and a structural premium for lottery-style payouts. Small flows move small markets, and the price floor tends to sit above the model.

Venue matters here too. Kalshi holds the sharpest prices across several title favorites, including Georgia, Ohio State, Notre Dame and LSU, while Polymarket carries the best price on Texas, Miami and Indiana. The cheapest venue is a per-team question, not a blanket one, and traders can compare venues before committing size (Kalshi code FADE, Polymarket code TGSWC).

When news moves a price, and when it is noise

The cleanest way to classify a move is to ask where it left the price relative to fair value. A move on real news should push price toward the probability the new information implies. A move on thin liquidity often overshoots, leaving price stranded above or below the model.

Texas is the current example of price leading the model. It prints 14c on Polymarket against a 12.2% fair value, the widest premium among the favorites. That gap is either the market pricing information the model has not yet absorbed, or a liquidity-driven overshoot that reverts. The model's read flags it as a fade candidate; it does not guarantee the price falls.

Miami shows the opposite alignment: 12c best price against 9.9% fair value, a smaller premium, while Indiana at 8c sits almost exactly on its 7.9% fair value. Prices resting on the model are doing what efficient markets should; prices floating above it are where the model sees an edge.

Using fair value as the anchor

Fair value turns a moving price into a decision. Once the model's probability is fixed, every price move can be scored: did it close the gap to fair value (informative) or widen it (a fade or a liquidity artifact)?

Across the board the pattern is consistent. Favorites with deep books, Georgia, Ohio State, Notre Dame, hug the model. Tail names, Ole Miss at 5c versus 3.3% and Oregon at 4c versus 4.4%, show the noisier fit that thin liquidity produces, with Oregon actually pricing a shade under its fair value.

None of this is financial advice, and the model can be wrong as easily as the market. But naming the force behind a move, liquidity, news or reversion to fair value, is what separates reading a price from chasing it.

The board at a glance

The two charts below use only current model figures: fair value for the top of the title board, and best available price in cents for a cross-section of favorites and tail names.

Fair value ranks the contenders; best price shows where each contract can be traded and how far it sits from the model's read.

National title fair value, top six
Texas12.2%
Georgia11.6%
Ohio State10.7%
Notre Dame10.7%
Miami9.9%
LSU8.2%

Best price in cents highlights the two ends of the board: favorites clustered near 12c to 14c, and tail contracts down at 4c to 5c where liquidity thins out.

Best price by team
Texas14c
Georgia12c
Ohio State12c
Miami12c
Ole Miss5c
Oregon4c
TeamsTexasGeorgiaOhio StMiamiOle MissOregon
Advertisement
Kalshi ad

Frequently asked questions

What makes prediction-market prices move?

Three forces drive most moves: liquidity (how much size the order book can absorb), news (injuries, results and roster changes that reset expectations), and mean reversion toward a model's fair value. On a given day one force usually dominates the other two.

How can a trader tell news from noise in a price move?

Compare the new price to the model's fair value. A move that closes the gap to fair value on real information reads as signal; a move that pushes price further past fair value on thin volume usually reads as noise or a liquidity air pocket.

Why do Kalshi and Polymarket show different CFB prices?

Liquidity concentrates differently by venue. Kalshi tends to carry the sharpest prices on national-title favorites, while Polymarket often holds the longer tail of contracts, so the cheapest venue shifts by team.

What does the model's fair value actually represent?

It is the model's estimate of the true probability a contract resolves yes, expressed as a percentage. A 12.2% fair value on Texas means the model reads roughly a one-in-eight chance, which maps to a fair price near 12c.

Does a price above fair value mean it will fall?

Not necessarily. Fair value is an estimate and both price and model can be wrong. A price above fair value flags a fade candidate, not a certainty, and thin markets can hold a gap for a long time.

About the author
The Model Desk

The Redshirt Analytics modeling team prices every college football futures contract and tracks the gaps between fair value and live market prices on Kalshi, Polymarket and ProphetX.