What does expected value tell us about possible outcomes?

Expected value is the probability-weighted mean of a random quantity. It describes the full distribution of possible outcomes rather than predicting which one will occur next. The word expected can be misleading if it is read as promised or most likely.

Consider an entirely imaginary token lottery with no real stakes. A draw changes a paper total by either −4 tokens or +8 tokens. In the initial model, each outcome has probability one-half. The expected change is one-half times −4 plus one-half times 8, which equals +2 tokens.

No individual draw produces +2. Each produces −4 or +8. A positive expected change therefore does not eliminate the possibility of a loss, and it does not guarantee a gain over a short sequence. The possible outcomes and their probabilities remain part of the description.

Changing the probability changes the mean without changing those two outcomes. If +8 has probability one-quarter, the expected change is −1 token. If it has probability three-quarters, the expected change is +5 tokens. These numbers summarize different distributions, not new single-draw results.

Over many suitably independent repetitions with stable probabilities, the average can approach the expected value. That long-run statement does not promise an exact total after a fixed number of draws. Nor does expected value alone decide whether risk is acceptable: variability, consequences and personal constraints are additional considerations.