Two traders each make a hundred bets over a year. The first wins 80 of them and loses only 20 — a stellar hit-rate any pundit would envy. The second loses 80 and wins just 20. At the office holiday party, who’s buying the drinks?
You can’t answer yet, and that’s the whole point. The first trader’s 80 wins each paid a dollar, and each of her 20 losses cost twenty — she’s down 60 dollars. The second trader’s 80 losses cost a dollar each, and each of his 20 wins paid twenty — he’s up 320 dollars. Same skill at picking winners? Irrelevant. The trader who was wrong four times as often finished more than five times richer, because his bets were lopsided the right way: small when he lost, huge when he won.
That lopsidedness has a name — asymmetry — and it is one of the most counterintuitive, most profitable ideas in this entire latticework. It quietly rearranges the question you ask before any risky decision. Not “how likely am I to be right?” but “when I’m right, how big is the win, and when I’m wrong, how big is the loss?” Get the shape right and you can be wrong most of the time and still win. Get it backwards and being right almost always won’t save you.
Before you read — take a guess
Trader A wins 80% of her bets; Trader B wins only 20% of his. Which one is more profitable over a year?
Two shapes: the smile and the frown
Every risky decision has a payoff shape — the picture of what you gain when it goes well against what you lose when it goes badly. Two shapes matter most, and they are mirror images.
A convex payoff is the smile: your downside is capped — a loss you can absorb and walk away from — while your upside is open-ended, with no ceiling on how good the good case can be. A small, known cost buys a shot at a large, unbounded gain. Buying a lottery-like startup stake for money you can afford to lose, running a cheap experiment that might discover something huge, planting ten seeds knowing nine will die — all convex. You lose small and often; you win big and rarely.
A concave payoff is the frown: the exact reverse. Your upside is capped — a small, steady gain — while your downside is open, with a rare catastrophe lurking underneath. You win a little almost every time, and once in a rare while you lose everything. Writing insurance on a disaster that “never” happens, over-leveraging to squeeze out a few extra percent, the strategy that prints money for years and then blows up in a week — all concave. Nassim Taleb’s phrase for it is unforgettable: picking up pennies in front of a steamroller.
The one-sentence version
A convex (smile) bet has a capped downside and an open-ended upside — lose a little often, win big rarely. A concave (frown) bet is its mirror: a capped upside and an open-ended downside — win a little often, lose big rarely. The shape of the payoff, not your hit-rate, decides where you end up.
The reason the two traders’ fortunes diverged so wildly is that they were standing on opposite shapes. Trader B held the smile; Trader A held the frown. And notice the cruelty of it: the frown feels wonderful right up until the steamroller, because you win and win and win. A concave strategy hands you a long, comfortable winning streak as it quietly loads the one loss that erases it. The convex strategy feels terrible — you lose, and lose, and lose — while it waits for the one win that pays for all of them. Comfort and correctness point in opposite directions.
Which of these is a CONVEX (smile) payoff — capped downside, open-ended upside?
Why this is more than a betting trick
If asymmetry were only about traders, it would be a finance footnote. It’s a genuine mental model because the shape transfers everywhere. A scientist running many cheap experiments is playing the smile: each one costs little, most fail, and the rare hit rewrites a field. A city building flood defenses to a level that has “never” been needed is refusing the frown. Sending a low-cost, exploratory email that might open a huge door; keeping several career paths alive instead of betting everything on one; trying a new skill on a weekend before committing a decade — all the same convex move. The model even tells you what to manufacture: you can deliberately engineer convexity into your decisions, which is exactly what optionality is, and what the back half of this course is about.
And it comes with a characteristic failure to watch for — one we’ll hunt down again and again: judging a decision by how often it works instead of how the payoffs are shaped. That single confusion is why people abandon good convex bets during their long losing streak, and why they pile into comfortable concave ones right up to the day the steamroller arrives.
A strategy has won money every single month for three straight years with almost no losing days. What should a student of asymmetry check FIRST?
The map of the course
Five teaching lessons, then a final exam you can’t undo. The route up:
- The Shape of the Payoff — the core arithmetic: value is probability times magnitude, and when magnitudes are lopsided, magnitude wins. We make convex and concave precise and you’ll drive an interactive payoff explorer — set the downside cap, the upside, and the win-probability, and watch the expected value move.
- Losing Most Bets, Winning the War — the convex bet worked out in full numbers (lose small on 80% of tries, win big on 20%, finish far ahead), against its mirror: the concave “pennies in front of a steamroller” trap, and why fat tails make the steamroller inevitable.
- Optionality — how to manufacture asymmetry: pay a small, known cost for the right but not the obligation to a big upside. Financial options, but also cheap experiments, renewable leases, and keeping paths open — plus why optionality loves volatility (the antifragile idea) and why it is never free.
- The Barbell & Jensen’s Inequality — the strategy that puts it to work: pair the very safe with the very risky and shun the fragile middle. Plus the one piece of math, stated in words: for a convex payoff, the average outcome beats the outcome of the average.
- Cut the Downside, Keep the Upside Open — the decision heuristic that ties it together: seek convex bets, run many small reversible experiments, let the winners run — and the traps to dodge (paying too much for optionality, mistaking a capped loss for an unbounded one, confusing mere variance with good asymmetry). Ends with a whole-course recap.
Then a Final Exam — graded, one question at a time, one-way: once you answer, it locks. No back button, no retries.
How to use this course
One habit does most of the work: before you judge a bet by how often it wins, draw its payoff shape. When you hit an exercise, commit to an answer in your head before revealing anything — the small sting of being wrong is what makes the idea stick. The exercises are the lesson; the prose just sets them up.
Next up: lesson 1, where we make the smile and the frown precise, put real numbers on “probability times magnitude,” and hand you the payoff explorer to drive yourself.