An average can describe almost nobody
Expected value is a legitimate and useful number, but it is a summary, and summaries lose information. When payouts are heavily lopsided, the average can sit at a value that almost no individual play ever produces. Treating that number as “what usually happens” is the average outcome trap.
Trap 1: A rare big prize inflates the average
Consider a game where you stake 1 coin. With probability 0.99 you lose it. With probability 0.01 you win 200 coins.
The expected net result per play is 0.99 × (−1) + 0.01 × 200 = −0.99 + 2.00 = +1.01 coins. On paper the game is generous.
Now look at what a player experiences. In a run of 100 plays, the chance of never hitting the prize is 0.99^100, which is about 0.37, or 37%. More than a third of the time, a player who followed the “positive” game would lose their entire 100 coins and see nothing in return. The average is carried by the rare event, and it only shows up reliably over a very large number of plays, far more than most people will ever make.
Trap 2: The reverse, small steady wins and a hidden disaster
Flip the structure. With probability 0.99 you win 1 coin, and with probability 0.01 you lose 100.
Expected net result: 0.99 × 1 + 0.01 × (−100) = 0.99 − 1.00 = −0.01. That looks almost harmless, and many sessions will show a steady stream of tiny wins. But a single loss wipes out about a hundred wins. The pattern of frequent small gains and occasional large losses can make a negative or neutral game feel safe. The numbers are the same kind of lopsidedness as the first trap, just pointing the other way.
Trap 3: Jackpot-style prizes and ticket price
Consider a ticket costing 2 that has a 1 in 1,000,000 chance of paying 1,500,000. The expected payout is 1,500,000 / 1,000,000 = 1.5, so the expected net result is 1.5 − 2 = −0.5 per ticket. The huge headline prize makes the ticket feel valuable, but in a typical experience the buyer receives nothing at all. Lottery odds work this way: nearly all the expected value lives in an outcome that almost nobody ever sees.
Trap 4: Ignoring what you can afford to lose
A positive expected value assumes you can keep playing until the long run arrives. If your resources run out first, you never get there. A player with 10 coins in the first trap’s game has a real chance of finishing before the 1-in-100 prize appears. This is the territory of risk of ruin: a good average does not help if one unlucky stretch ends the game. In a game where a bad outcome cannot be undone, such as losing a unique character or money you need, the average is the wrong thing to optimise.
Trap 5: Mixing different players into one average
Pooled averages can hide how different players fare. If a few players find a very rare item and everyone else finds nothing, the average value of finds per player looks decent, yet most players are empty-handed. The same applies to published win rates and loot statistics: a single mean from a skewed pool says little about a typical case.
How to avoid the traps
Several habits help.
- Ask what the typical outcome is. The middle result, or the chance of any win at all, often tells a different story than the mean. The gap between them is the subject of the median outcome discussion elsewhere on this site.
- Look at the full list of outcomes. Write out each result and its probability, not just the weighted total. A short table usually exposes the skew immediately.
- Estimate the chance of a bad stretch. For the 1% prize game, the chance of 100 straight misses is easy to compute and sobering.
- Check your limits. Decide what you can lose without harm before looking at averages, not afterward.
- Be careful with emotional framing. A large possible prize can push people to overweight it, which is partly why long shots feel more attractive than their probabilities justify. Related patterns are described under cognitive biases.
What to take away
Expected value is a fair measure of the long-run average, and a poor measure of a single player’s likely experience when payouts are skewed. In games that involve real money, a positive-looking average does not guarantee that you will win, and a high headline prize does not make a negative-value game profitable. If gambling is causing stress or loss, support is available, and treating any game as a source of income is a risk in itself.