A 92% win rate can still lose you money

The seduction of the win rate

Selling an out-of-the-money option feels wonderful most of the time. You collect cash up front, the stock wanders around, the option expires worthless, and you keep the credit.

Do it ten times and you might win nine. Every options-income screen on the internet leads with that number — the probability of profit, or POP — because it is the number that feels best. It is also, on its own, almost meaningless.

Probability of profit tells you how often you win. It says nothing whatsoever about how much you lose when you lose.

A coin flip you should refuse

Imagine a bet that pays you $50 nine times out of ten, and costs you $800 the tenth time. Your win rate is a superb 90%. Here is the full ledger for ten plays:

Played repeatedly, this bet reliably drains your account — while feeling great almost the whole time. Ninety percent of your statements say you were right. That is the exact shape of a badly priced short option.

The mistake most sellers make is not selling options. It is selling options priced so that the rare loss is bigger than all the wins put together, and never noticing, because the win rate looked superb the whole way down.

Why the payoff is built this way

This isn’t bad luck; it is the structure of the instrument. When you sell an option, your profit is capped at the credit you collected the moment you sold it. Your loss is not capped at anything.

Look at what POP actually measures on that chart: the chance you land anywhere in the green region. It treats a $1 profit and a $2,000 loss as the same kind of event — one is a win, one is a loss — and reports only the count. Any strategy whose losses live in an unbounded region needs a second number.

The second number: expected value

For every candidate trade, the engine models the whole distribution of outcomes — not just the odds of winning, but the size of the losses out in the tail, based on how the underlying stock actually moves rather than how the option market says it might. From that it computes the trade’s expected value: the average of every outcome, weighted by how likely it is.

If that average is not positive, the trade is rejected. It does not matter how high its win rate is. This is the one gate that has no override, because a negative-expectation trade with a high win rate is the single most expensive thing an income seller can own — it pays you for months and takes it all back in an afternoon.

How to read this on your own screen

Alongside the win rate, every opportunity shows its modeled expected value and its tail risk — roughly, “when this trade loses, how bad is the average loss?” Read all three together, in this order:

1. Expected value first. If it isn’t positive, nothing else on the row matters.
2. Tail risk second. Ask whether you could take that loss twice in a month without changing your behaviour. If not, the position is too large, whatever the win rate says.
3. Win rate last. It tells you what the ride will feel like, not where it ends.

This is also why the app will sometimes refuse a trade whose POP looks spectacular. A 92% winner that models to a negative expectation is not a high-probability income trade; it is a slow leak with good manners. The Methodology page explains how these numbers are computed and how we check them against history.