Scoreboard: insiders vs the crowd
Every prediction market we track eventually resolves yes or no — and once it does, both sides of this site's premise can be graded. Below: how well-calibrated the crowd's final odds were across everything that has resolved, and — for questions with a clear bullish or bearish reading — whether the insiders or the crowd called it. Updated nightly; every call links to its market.
Was the crowd's confidence honest?
For every resolved market, we take the crowd's final odds before resolution and ask how often "80% likely" things actually happened. A perfectly calibrated crowd tracks the shaded target exactly.
Head-to-head calls
Only markets where yes/no has an unambiguous bullish or bearish meaning for the stock qualify — earnings beats and high/low price targets. The insider stance is the direction of net discretionary insider flow while the market was live (10b5-1 plans excluded). Disagreements — where insiders leaned one way and the crowd the other — are listed first: they are the whole point of this site.
What's excluded, and why
Two honest caveats. The head-to-head sample is small and will stay small for a while — most tracked markets are KPI ranges ("operating margin 34–36%?") where an insider trade implies nothing about yes or no, and many qualifying markets had no meaningful insider activity to grade. And an insider sell is a much weaker signal than a buy (people sell for taxes and houses; they buy for one reason) — grading sells and buys equally here is deliberate and conservative. We publish the record either way, as it accumulates, wins and losses alike.
Not investment advice. Odds are each market's last stored daily snapshot before resolution; insider figures come from SEC Form 4 filings.