The Asymmetry of Stakes / Decision-Making

The shape, not the odds

Under risk you manage probabilities. Under uncertainty, where the probabilities are fog, you manage something you can still see clearly: the shape of the payoff — how much you lose if it goes wrong, how much you gain if it goes right.

This is liberating, because the payoff shape is often knowable even when the odds are not. You may have no idea whether a venture will succeed. You can still know that failing costs you a fixed, survivable amount and that succeeding pays an open-ended one. That asymmetry is actionable without any forecast at all.

Convex and concave

A bet is convex when the downside is bounded and the upside is large or unbounded: small fixed loss, big open gain. A bet is concave when it inverts: a small or capped gain sitting over a large, sometimes catastrophic, loss.

The two demand opposite policies under uncertainty. Convex bets are worth taking even if they usually fail, because the rare success pays for all the cheap failures and then some. Concave bets are worth refusing even if they usually work, because the rare catastrophe erases all the small wins and then your position. The probability barely matters; the shape decides.

Under uncertainty, prefer bets where being wrong is cheap and being right is large. Refuse bets where being right is small and being wrong is ruinous — no matter how unlikely the ruin looks.

Optionality

The convex shape has a name in its purest form: an option — the right, without the obligation, to take something if it turns out well. You pay a small premium; you keep the upside; you can walk away from the downside. An option converts uncertainty from a threat into a resource, because the more volatile and unpredictable the world, the more a capped-loss, open-gain position is worth.

This reframes how to act when you cannot predict. You do not need to know which experiment, which relationship, which idea, which venture will pay. You need to take many convex positions — each cheap to lose, each open to win — and let the unpredictable world select the winners. The strategy does not require foresight. It requires arranging your exposures so that foresight is unnecessary.

The error in both directions

Most decision failures under uncertainty are failures of shape, and they come in two forms.

The first is taking concave bets for their high probability of a small win: picking up pennies in front of a steamroller, accumulating modest reliable gains while quietly carrying an exposure that wipes them all out the one time it triggers. It looks prudent — it usually works — right up until it doesn’t, once.

The second is refusing convex bets for their high probability of a small loss: declining cheap experiments because most of them fail, and thereby forgoing the rare, oversized payoff that the whole portfolio of failures was buying. It looks disciplined. It is the slow starvation of anyone who will not pay small certain costs for large uncertain gains.

The limit of the shape

But there is a feature of a payoff that the convex/concave picture, on its own, does not capture — and it is the most important feature of all. “Bounded downside” quietly assumes you survive the downside to keep playing. Some losses are not like that. Some are final.

A bet can have a beautiful shape on paper and still be uniquely deadly, because losing it removes you from the game entirely — no re-roll, no recovery, no next bet over which the average was supposed to work out. Before any talk of payoff shapes, one distinction sits underneath them all: whether a decision can be undone, or whether it is a door that locks behind you. That is the next essay.