
Hype Test
Before you bet on, write about or teach a Polymarket or Kalshi market, Hype Test checks whether the price matches the public evidence, and shows its reasoning lens by lens so you can judge it yourself.
- Who it helps
- Forecasters, journalists, teachers, students, careful bettors.
- Skill it shows
- Designing an AI reasoning tool, then testing it in public and reporting the misses.
Proof
First public run, on 20 races in the 2026 midterms, scored with no retrospective edits. Called 16 of 19 decided races; rated Nevada governor a 50/50 toss-up. The three misses (Iowa Senate, Maine's 2nd, Texas governor) stay on the scorecard with what they taught. A comparison with the markets, not a contest.
How it works
Five lenses look at the same event: fallacies and distortions in the coverage; testing the resolution (if it were true, what else would have to be true?); hype versus reality; trajectory (where the trend is heading by the deadline); and shifts in language over time. Four of the five never see the market price. A composite score shows where the lenses agree.











