Every probability, scored against the world.
We publish the Brier score for every Scenario Question that has resolved on the platform. Per scenario. Per horizon. Public, with no login. This is the most defensible artifact the engine ships — and the longer it runs, the harder it is to fake. Compare us to Metaculus, Polymarket, GPT-class baselines, or Tetlock's Superforecasters at the bottom of the page.
What we measure
Calibration curve
0 scored resolutions · 10 binsWhen we say 70%, does it happen 70% of the time? Each dot is a probability band; its height is how often those calls resolved yes. Perfect calibration sits on the diagonal. (A Brier score is the squared error of the probability — 0 is perfect, 0.25 is the coin-toss line for binary questions.)
No questions have reached their horizons yet. Every live question is already committed to this chart — each resolution lands here as a point, on the record, hash-chained.