Resolution criteria

This market will resolve based on which prediction platform demonstrates the highest accuracy for the 2026 US midterm elections.

Each answer (platform) will resolve to YES if it is determined to be the most accurate according to the criteria below, and NO otherwise. In the event of a tie, multiple platforms may resolve to YES. Any platform that does not meet the criteria or is missing sufficient data will resolve to NO.

The market will resolve according to the following hierarchy:

1. Primary Source: If a reputable comparative analysis, tracker, or academic study (such as Brier.fyi or an industry-accepted post-election report) publishes comparative Brier scores or relative accuracy metrics evaluating multiple prediction markets specifically for the 2026 US midterm election cycle, the platform(s) with the best score (e.g., the lowest Brier score) in that report will resolve as YES.

2. Fallback Metric (House & Senate Control Brier Score): If no comprehensive comparative study is available by June 1, 2027, accuracy will be calculated using a standard Brier score on the two primary outcomes of the midterms: US Senate Control (which party wins the majority) and US House of Representatives Control (which party wins the majority):

   • Formula: $BS = \frac{(1 - p_{\text{Senate}})^2 + (1 - p_{\text{House}})^2}{2}$

   • Where $p_{\text{Senate}}$ and $p_{\text{House}}$ represent the platform's implied probability (expressed as a decimal from 0 to 1) for the correct eventual outcome, sampled at 11:59 PM Eastern Time on November 2, 2026 (the night before Election Day).

   • The platform(s) with the lowest Brier score will resolve YES.

   • Platforms that did not host liquid, publicly accessible markets for both of these outcomes will resolve NO under this fallback.

Background

The 2026 US midterm elections are a major test for the prediction market ecosystem. Since the CFTC-related regulatory shifts and expansion of platforms like Kalshi alongside established venues like Polymarket, PredictIt, Manifold, and Metaculus, competition over forecasting accuracy has intensified.

Accuracy is typically evaluated using the Brier score, a proper scoring rule that measures the calibration of probabilistic forecasts. A Brier score of 0 represents perfect prediction, while 0.25 represents a permanent coin-flip. This market assesses which platform's crowd wisdom is best calibrated when tested against the actual election outcomes.