Unit test scenarios and Monte Carlo simulation (500 trades). Demonstrates LMSR stability and max-loss bound.
4cast uses the same Logarithmic Market Scoring Rule (LMSR) and calibration as in these runs: when you buy or sell shares in a market, prices and outcome probabilities come from this cost function. This page replays that logic on synthetic trades to check that probabilities stay valid, total exposure never exceeds the max-loss bound, and no invariants are violated. This way, the live markets behave as expected. Read more about this here.
Select scenario and outcomes, then click Run to run unit test and 500-trade Monte Carlo.
This section shows a toy game to illustrate how expected value (EV) and a small house edge work. It is not using the exact odds or payouts of the real games on 4cast; it is only an example to build intuition.
If you played this game many times, on average you would lose 0.100 coins per play.
When the expected value is slightly negative for the player, the game has a small house edge. Real games on 4cast are designed with a modest house edge in a similar spirit to this example, but their exact odds and expected value are not shown here.