Sports made up about 85% of 2,681 resolved Limitless contracts, limiting what the analysis says about whether prices can track outcomes in economic or political markets. Tiger Research published the analysis Thursday, Oct. 22, using public Limitless API records to compare quotes from the day before settlement with recorded outcomes.
The share settling yes came within four percentage points of the quoted odds in nine of the 10 probability bands. Among 113 contracts priced at 90% or more, 97.3% settled yes; in the 70% to 80% band, 74.5% of 165 contracts did.
In a yes-or-no contract, the quote implies a chance of the event occurring, with a $1 payout if it happens and $0 if it does not. Calibration checks whether contracts with similar quotes settle yes at similar rates across a group. It describes group performance, not the outcome of any individual contract.
In the 50% to 60% band, 48.8% of contracts settled yes. About one-third, mostly smaller sports markets, still had their original 50% quote the day before settlement. The analysis linked the larger discrepancy in that band mainly to thin trading.
Limitless partly funded the analysis. Its published findings did not include the full contract list or enough detail on the API queries for outsiders to reproduce the calculations, limiting review of the results.
A separate Federal Reserve Board working paper examined U.S. interest-rate forecasts from Kalshi. For its sample beginning in 2022, the median and most frequently selected rate estimates averaged no distance from the eventual target when measured on the eve of Federal Open Market Committee (FOMC) meetings. The paper’s authors also reported a statistically significant difference compared with federal funds futures, a market benchmark for rate expectations. Its analysis attributed much of the gap to the Fed’s September 2024 decision to cut rates by 50 basis points: Kalshi markets had leaned more heavily toward that outcome. The authors called the paper preliminary and said it does not state the Federal Reserve Board’s position.
The findings add to EconoTimes’ earlier coverage of a review of collective forecasts across 12 domains. The Limitless and Kalshi research test different markets and questions, so their results remain separate measures rather than a single score for prediction-market accuracy.
