Verified & closed
Events already resolved. The calibration and reputation base for everyone who took a position. Value decays over time but never reaches zero — it's the epistemic backing.
value = f(age, citations) · floor > 0Public record of claims · event observatory
Ask what's known, what's happening, and what's likely — with the why.
A continuous, auditable record of claims about world events, in three layers that feed each other: verified past, reported present, priced future — every price and every verified fact carries the argument that backs it.
Live today: weather reports & forecasts — the record's first shipped layer.
How the record is organised
Three temporal layers that feed each other
A prediction becomes a reported event, which becomes a verified record — and the outcome recalibrates whoever forecast or reported it.
Being verified now
Reports coming in about ongoing events. Priced by novelty, verifiability, scarcity of coverage, and active demand. The first report of an event is worth far more than the fiftieth.
price = novelty × scarcity × demandPriced by the market
Predictions about events that haven't happened. A market maker sets the price, and the price is the implied probability — readable directly, no interpretation needed.
price = implied probability (LMSR)The argument graph runs across all three layers — connecting claims through support and attack edges, so every price and every verified fact has a visible reasoning trail, not just a number. When a premise falls in the present layer, the markets that rested on it are flagged automatically.
Live now
The first layer is already live: weather
Before the argument graph and the markets, the record starts with something concrete: real-time weather forecasts, conditions reported by people on the ground, and alerts — the same present-layer mechanics (novelty, verification, active demand) the rest of the product is built on.
What EntropyCast gives you
For anyone deciding something
Calibrated probability, with a track record attached
Not "experts say X might happen." Instead: 68%, up from 40% yesterday because a premise fell — and whoever is moving that price has a public, auditable forecasting history you can inspect.
Transparency
The mechanism is real, and it's explained
Prices are set by a market maker; contribution value is a formula; the platform's own health is kept in range by an automated control loop. None of that is hidden — it's one click away whenever you want it, and out of the way when you don't.
Some control-loop parameters and response timings aren't published — deliberately, to stop them being gamed. The objectives are public; the exact coefficients are not.