Decision-making and strategy
Asymmetric bets
Take bets where the most you can lose is small and known and the most you can win is large, even if you expect to lose most of them.
No single originator: the idea runs through gambling and finance. Jeff Bezos stated it plainly in his 2015 letter to Amazon shareholders: "Given a ten percent chance of a 100 times payoff, you should take that bet every time."
Judge a bet by its payoffs weighted by their odds, not by how often it wins. Bezos's example is a bet with a 10 percent chance of paying 100 times the stake: on average it returns ten times what you put in, yet you will be wrong nine times out of ten. He contrasted business with baseball, where the best possible swing scores four runs; in business, now and then, one swing scores 1,000.
When losses are capped and gains are not, a few winners pay for many losers. Most people avoid such bets because losing feels bad and losing often feels worse. That is exactly why they are underpriced. Make enough of them and the long tail of the distribution, not your hit rate, decides how you do.
It only works if you can survive the losing streak and place enough bets for the odds to show. The odds are usually unknown, and people flatter their own chances: a lottery ticket has a capped loss and a huge prize, and is still a bad bet. Check for hidden downside too. A bet that looks cheap can carry legal, reputational or ruinous risk, and you mostly hear about the winners, not the many who placed the same bets and lost.
Example
In 2004 Peter Thiel invested $500,000 in Facebook, then a site for college students. The most he could lose was that $500,000. When Facebook went public in May 2012 he sold 16.8 million shares for $640 million. After the lock-up period ended that August, he and his funds sold about 20 million more for $395.8 million, at between $19.27 and $20.69 a share, taking his total proceeds to just over $1 billion according to CNNMoney.
Where it applies
Spend a day testing an idea that is unlikely to work but would change the product if it did. Try a new open model against your evaluation set; the cost is an afternoon and the upside is a cheaper or better system. Apply for the stretch role, write the post, send the email to someone you admire. Keep each bet small enough that losing it changes nothing important.
Think about it
What cheap bet with a large possible upside have you been avoiding only because it will probably fail?
Applied it to something real? Log it.
Test yourselfHow can a bet you expect to lose nine times out of ten still be a good one?
If the loss is small and capped while the rare win is large enough, the expected payoff is positive. Bezos's ten percent chance of a 100 times payoff returns ten times the stake on average.
Related models
- OptionalityPrefer positions that give you the right, but not the duty, to act later, so you can take the upside when it appears and walk away when it does not.
- Probabilistic thinkingThink in odds rather than certainties: estimate how likely each outcome is, and judge a decision by those odds rather than by the single result that happened.
- Margin of safetyLeave a wide gap between what you expect to need and what you build or pay, so that your errors and bad luck do not sink you.
- Loss aversionA loss hurts more than an equal gain pleases, so people pay too much to avoid losses and take odd risks to escape them.
- Moral hazardWhen someone else will bear the cost of your risks, you take more of them.