Prediction Market Statistics
Updated October 2026
16 one-sentence findings about the markets themselves — how often the market favorite wins, how prices line up with outcomes, what trading costs, and how Kalshi and Polymarket US compare. Every number is computed from our archives each day, with its sample size and as-of date, and every finding has its own link.
Licence and citation
Free to quote and republish under Creative Commons Attribution 4.0 (CC BY 4.0). Cite as “PredictionMarketsPicks” and link the finding.
Example: Source: PredictionMarketsPicks, predictionmarketspicks.com/data/prediction-market-statistics#fed-1
Market favorites
In the 2026 NFL regular season through October 26, 2026, the Kalshi market favorite at the last archived price before kickoff won 66 of 104 games (63.5%), at an average favorite price of 65.1¢.
- As of:
- October 26, 2026
- Sample:
- 104 games, September 9, 2026 – October 26, 2026
- Source:
- Kalshi KXNFLGAME price archive × official NFL finals
In the 2026–27 NHL regular season through October 26, 2026, the Kalshi market favorite at the last archived price before puck drop won 114 of 187 games (61%), at an average favorite price of 58.8¢.
- As of:
- October 26, 2026
- Sample:
- 187 games, September 29, 2026 – October 26, 2026
- Source:
- Kalshi KXNHLGAME price archive × official NHL finals
Calibration: do favorites win as often as their price says?
Across 156 NFL and NHL games through October 26, 2026 in which the Kalshi favorite was priced 50–59¢ before the start (average 54.9¢), the favorite won 54.5% of the time (85 of 156).
- As of:
- October 26, 2026
- Sample:
- 156 games
- Source:
- Kalshi game-contract archive × official finals (NFL 2026, NHL 2026–27)
Across 92 NFL and NHL games through October 26, 2026 in which the Kalshi favorite was priced 60–69¢ before the start (average 64.3¢), the favorite won 68.5% of the time (63 of 92).
- As of:
- October 26, 2026
- Sample:
- 92 games
- Source:
- Kalshi game-contract archive × official finals (NFL 2026, NHL 2026–27)
Across 35 NFL and NHL games through October 26, 2026 in which the Kalshi favorite was priced 70–79¢ before the start (average 74.3¢), the favorite won 80% of the time (28 of 35).
- As of:
- October 26, 2026
- Sample:
- 35 games
- Source:
- Kalshi game-contract archive × official finals (NFL 2026, NHL 2026–27)
Cross-venue price gaps
From October 22, 2026 to October 27, 2026, our sports scanner logged 644 Kalshi vs Polymarket US price gaps across 230 markets; the median gap was 0.98¢ per contract, 3.9% were 3¢ or more, and Polymarket US was the cheaper venue in 73.6% of them.
Polymarket US prices as of October 27, 2026. Sponsored by Polymarket US.
- As of:
- October 27, 2026
- Sample:
- 644 logged gaps, 230 markets
- Source:
- Sports price-gap log (Kalshi vs Polymarket US order book)
From October 2, 2026 to October 27, 2026, our sports scanner logged 1,317 Kalshi vs Polymarket.com (international order book) price gaps across 637 markets, with a median gap of 1.71¢ per contract and 10.5% at 3¢ or more.
- As of:
- October 27, 2026
- Sample:
- 1,317 logged gaps, 637 markets
- Source:
- Sports price-gap log (Kalshi vs the Polymarket.com international book)
The Fed
Kalshi's favorite on the last day before the decision won 14 of 14 FOMC rate decisions from January 29, 2025 to September 16, 2026, at an average price of 95.5% on the outcome that happened; Polymarket.com's favorite won 14 of 14.
- As of:
- September 16, 2026
- Sample:
- 14 decisions
- Source:
- Daily closes per venue on the realized outcome (fed_venue_daily, per-meeting basis)
The closest call in that record was the Jul 28–29, 2026 decision (hold), which Kalshi priced at 74.5% and Polymarket.com at 76.9% the day before.
- As of:
- September 16, 2026
- Sample:
- lowest pre-decision price across 14 decisions
- Source:
- Daily closes per venue on the realized outcome (fed_venue_daily, per-meeting basis)
Before the Sep 15–16, 2026 FOMC decision (25bp hike), Kalshi's price on that outcome first closed above 50% on August 28, 2026, 19 days before the decision, and closed at 86.5% the day before; it never closed at 90% or higher (32 daily closes from August 15, 2026).
- As of:
- September 16, 2026
- Sample:
- 32 daily closes
- Source:
- Kalshi daily closes on the realized outcome (fed_venue_daily, per-meeting basis)
2026 midterms
Since Labor Day (September 7, 2026), Kalshi's implied probability that Democrats control the Senate after the 2026 midterms ranged from 48% (September 10, 2026) to 68% (October 23, 2026) across 51 daily closes, and stood at 63% on October 27, 2026.
- As of:
- October 27, 2026
- Sample:
- 51 daily closes
- Source:
- Kalshi Congress balance-of-power market (four legs, normalized)
As of October 27, 2026, we track a Kalshi market for 35 of the 35 Senate seats on the 2026 ballot and for 66 individual competitive House districts.
- As of:
- October 27, 2026
- Sample:
- 35 Senate + 66 House race definitions
- Source:
- The race list behind /midterms-2026
Fees
Kalshi's taker fee on one contract works out to 0.63¢ at a 10¢ price, 1.75¢ at 50¢ and 0.63¢ at 90¢, because the fee is 7% × price × (1 − price), charged at the fill whether the position wins or loses.
- As of:
- September 23, 2026
- Sample:
- the published fee formula
- Source:
- Kalshi fee schedule, read in a browser
Because Kalshi rounds each fill up to the next cent, one 50¢ contract pays a 2¢ fee while 100 contracts at 50¢ pay $1.75 in total, or 1.75¢ each.
- As of:
- September 23, 2026
- Sample:
- the published fee formula
- Source:
- Kalshi fee schedule, read in a browser
Coverage
On October 26, 2026, our daily archive recorded closing prices for 419 Kalshi contracts across 5 categories (sports 294, economics 46, politics 38, crypto 30, weather 11).
- As of:
- October 26, 2026
- Sample:
- 419 contracts
- Source:
- Daily Kalshi close archive
On October 27, 2026, our politics archive recorded a closing price for 351 Kalshi election contracts.
- As of:
- October 27, 2026
- Sample:
- 351 contracts
- Source:
- Daily Kalshi politics archive
Method
Market favorites
For every finished regular-season game we take the last price we archived for each team’s Kalshi game contract strictly before the scheduled start (NFL: within 36 hours of kickoff, usually the game-day morning pass; NHL: within 2 hours of puck drop, captured about every 10 minutes). The price is the mid of a two-sided book no wider than 15¢, otherwise the last trade. The favorite is the side above 50¢; games at exactly 50¢ and ties are left out. The winner comes from the official final score. These are archived prices, not the exact closing tick.
Calibration
The same NFL and NHL games, pooled and grouped by the favorite’s pre-start price. A band is shown only once it holds at least 20 games, so higher bands appear as the season fills in. A well-calibrated market’s favorites win about as often as their average price says.
Cross-venue price gaps
Our sports scanner compares each Kalshi game contract with the same contract on Polymarket US and on the Polymarket.com international book every 15 minutes, before the game starts, and logs the first rankable gap per market, per venue, per day. A gap is the per-contract saving at the cheaper venue, fees included. These are gaps the scanner logged, not every matched market it checked. Polymarket.com and Polymarket US are different venues and do not carry the same price.
The Fed
Each venue is graded on its close on the last day before the decision, on the outcome that happened, comparing per-meeting markets only. Decisions before August 30, 2026 are reconstructed from each venue’s own public price archive; daily three-venue capture runs from that date. The run-up finding uses the most recent decided 2026 meeting with at least two weeks of daily Kalshi closes. The full record is on the Fed decision record.
2026 midterms
Senate control is read from Kalshi’s four-way Congress balance-of-power market: the two legs in which Democrats hold the Senate, divided by the sum of all four legs, from each day’s archived close. Labor Day is computed as the first Monday of September. Race counts come from the race list that drives /midterms-2026.
Fees
Computed from Kalshi’s published taker formula, 7% × price × (1 − price) per contract, rounded up to the cent on each fill; schedule last read September 23, 2026 at kalshi.com/fee-schedule. Resting (maker) orders pay no fee on most standard markets.
Coverage
Counts of distinct contracts with a closing price stored in our daily archives on the date shown. The general archive uses the last complete day, since the current day may still be filling.
What is deliberately not here: our own pick record or model output, anything derived from licensed data, and any number we could not compute from a stored record. Figures refresh daily; when a figure changes, the sentence changes with it, and the as-of date says which reading you are quoting.
More: the open data catalog (JSON + CSV) · Learn hub · methodology