Cheap Football Contracts: 0.6% Hit a 1.5% Price
Delta Arc's football backtests show cheap contracts are priced close to right — the spread and fee are where the cents go.
Published 2026-10-20 · Data as of 2026-10-02 · Market & data intelligence · Educational, not advice.
Delta Arc backtested 33,347 settled Kalshi football markets and found prices well calibrated at the mid, even on low-priced contracts. The catch: you buy at the ask, not the mid, and the fee bites hardest where contracts are cheap. Real patterns show up — rare outcomes run a touch rich — but they are too small and capital-heavy to trade.
A low-priced contract looks like a small, cheap ticket on an unlikely outcome. Three cents to win a dollar. On a prediction market, that three cents is not a price a house set for you. It is where the last buyer and the last seller agreed, on an exchange that matches the two of them and takes a fee, and never takes the other side itself. The number moves continuously until the outcome is known.
So the honest question about cheap contracts is not whether they pay off rarely — of course they do — but whether the price is right about how rarely. Delta Arc has been measuring exactly that across football markets, and the answer is more interesting than cheap stuff is a trap.
What the backtest actually found
Delta Arc's research (record DAKOS-BM-0031) backtested 33,347 settled Kalshi NFL and college football markets from the 2026 season through October 2, 2026. Priced at the midpoint between bid and ask, Kalshi came out well calibrated in both leagues — meaning the mid price, on average, tracked how often those outcomes actually happened.
That calibration holds even at the bottom of the board. In the college football set, contracts priced 0 to 3 cents settled YES 0.6 percent of the time, against an implied 1.5 percent at the mid. Small gap, but the direction matters: the realized rate sat a touch below the price. The market was not wildly wrong about rare outcomes. It was close, and slightly rich.
Why the mid flatters the buyer
Here is the catch that makes well calibrated something other than profitable to buy. Calibration is measured at the mid. You do not trade at the mid. A buyer pays the ask, which sits above the mid, and then pays the exchange fee on top.
Delta Arc's same college study found that a buyer paying the ask lost money in essentially every price bucket, on both the YES and the NO side, by 1 to 6 cents per contract after the exchange fee. Not because the market was badly priced, but because the spread plus the fee is a real cost, and on a cheap contract that cost is large relative to the few cents at stake.
Kalshi's fee schedule, effective July 7, 2026, charges 100 contracts $0.63 in fees at a 10-cent price, $1.75 at 50 cents and $0.63 at 90 cents. Delta Arc's arithmetic on those figures puts the fee at roughly 6.3 percent of the money paid at 10 cents, 3.5 percent at 50 cents and 0.7 percent at 90 cents. The fee is U-shaped by design, and it weighs most exactly where contracts are cheap.
NFL props: a sharper version of the lesson
The NFL data (record DAKOS-BM-0032) sharpens it. Across 48 regular-season games through October 2, 2026, buying contracts priced 25 cents or less at the ask returned +0.5 cents per contract, with a t-statistic of 0.27 — statistically indistinguishable from breaking even. Nothing there to act on either way.
Go cheaper and the picture shifts. On first-touchdown-scorer contracts priced 6 cents or less, buying YES at the ask lost 2.3 cents per contract, a t-statistic of minus 3.32, over those same 48 games. The identical contracts priced above 6 cents came out at +0.3 cents, roughly fair. The realized rates say why: anytime-touchdown contracts at 0 to 3 cents hit 1.2 percent against a 1.3 percent mid, and first-touchdown contracts at 3 to 6 cents hit 2.8 percent against a 4.2 percent mid. Rushing-yards overs priced 3 to 25 cents landed 1 to 2 percentage points under their mid.
Small and real is not the same as tradeable
An earlier study of the 2025 season (record DAKOS-BM-0003) saw the same shape. Across 2,976 anytime-touchdown markets, confirmed on 2,030 first-touchdown markets, touchdown contracts priced 5 to 15 cents hit 7.5 to 8.4 percent of the time against roughly 10 percent implied by price. Consistently a little rich.
But taking the other side meant buying NO at about 90 cents to earn roughly $0.006 to $0.020 per contract. Delta Arc judged that too small and too capital-heavy to be tradeable — you tie up 90 cents to chase a fraction of one. A measurable pattern is not the same as a usable one.
One more trap worth naming. An automated test that bought 2026 football contracts priced 10 to 20 cents at least seven days before kickoff closed 141 positions through October 2, winning 5.7 percent and averaging minus $0.150 per contract after fees; 133 of the 141 were college games, and the rule was wound down. The tell: 110 of those 141 actually filled at 20 cents or more, because the rule read the mid while orders filled at the higher ask. The spread alone pushed a cheap contract out of its intended range.
How we see it, and where to look
Delta Arc priced each market from the last hourly candlestick before kickoff, using Kalshi's API, which serves price history as 1-minute, 1-hour and 1-day candlesticks with bid, ask, volume and open interest.
Delta Arc's Prediction Markets product shows the live board across Kalshi and Polymarket in one place, so the mid, the spread and the fee sit together rather than one at a time. The lesson from the data is quiet and durable: on cheap contracts, the price is usually close to honest, and the spread and fee are where the cents go. The next question we are chasing is whether that gap ever opens wide enough, on any venue, to matter — and what it looks like on the tape when it does.
Sources: Delta Arc research, Kalshi's fee schedule, Kalshi API docs.
Sources
Outside pages cited in this post, in order of first citation
- Kalshi's fee schedulekalshi.com
- Kalshi's APIdocs.kalshi.com
This is the free read. Delta Arc Prediction Markets shows you every top market across Kalshi and Polymarket in one live view. Get early access.