Expected Value

A trade with a 30% chance of making $300 and a 70% chance of losing $100 has an expected value of +$20. Most people reject this trade because the loss is more likely than the gain. But run it 1,000 times and the expected outcome is +$20,000. Expected value does not describe any single outcome — it describes the average across many repetitions.

Every strategy the botwir3 builder produces is a system of repeated decisions. Each decision has a distribution of possible outcomes. The strategy is profitable if the expected value of its decisions is positive after accounting for spreads, fees, and slippage. It is unprofitable if the expected value is negative. The win rate alone does not determine profitability — a strategy that wins 30% of the time but makes 5x on each win and loses 1x on each loss is highly profitable.

The Kelly criterion, the position sizing logic, and the tolerance band all operate on the assumption that the strategy has positive expected value. If it does, these tools optimize how much to risk and how much deviation to tolerate. If it does not, these tools manage the rate at which money is lost — they do not prevent it. The gate validates discipline. It does not validate that the strategy's expected value is positive. Determining that is the user's responsibility, through backtesting, analysis, and ongoing monitoring.