Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts

The summary

Twenty-six seconds left in Super Bowl XLIX, second down on the one-yard line, and Seattle’s Pete Carroll called a pass instead of handing to Marshawn Lynch. It was intercepted; USA Today called it the worst play call in NFL history. Yet none of the sixty-six passes thrown from an opponent’s one-yard line that season had been intercepted, and the fifteen-year rate was 2 percent. A defensible call, a disastrous result. Poker players have a name for what his critics did: resulting — grading a decision by how it turned out. Two things decide how a life goes, decision quality and luck, and most of our errors come from refusing to separate them.

Life is poker, not chess

John von Neumann built game theory on poker, not chess. Real life, he said, “consists of bluffing, of little tactics of deception, of asking yourself what is the other man going to think I mean to do.” Lose at chess and there was a move you missed. Life is the other game: you can run a red light and sail through, or obey every signal and get hit anyway.

Ask executives for their worst decision of the past year and nobody names a well-reasoned choice that flopped. One CEO named firing his president, because sales fell afterward — never mind the benchmarking against competitors, the coach hired for the man, the alternative of splitting the role that was weighed and rejected. Good process, bad result, filed as a mistake. He decided he drove better drunk.

Every decision is a bet on an uncertain future

In the 1990s a Vegas card game talked high-stakes pro John Hennigan into moving to Des Moines for a month for $30,000, confined to one street where everything closed at 10 p.m. Two days in, he called to settle, opening with an offer that they pay him $15,000. Days later he paid them $15,000 to come home.

Strip off the folklore and the analysis is what anyone runs on a job offer: upside, downside, opportunity cost. A bet is “a choice made by thinking about what will probably happen,” which makes parenting, hiring and ordering the chicken bets too. Most aren’t against another person — they’re against every future version of you that you didn’t pick.

Say “I’m not sure,” then put a number on it

Your bets are only as good as your beliefs, and beliefs form carelessly. We assume we hear something, vet it, then believe it; Daniel Gilbert’s experiments found the real order is hear, believe, and only sometimes vet later. Hence the habit of multiplying dog years by seven — a ratio someone made up around the thirteenth century.

Intelligence doesn’t rescue you. Dan Kahan gave people data about a skin cream and the numerate ones read it correctly. Relabel the same numbers as gun control and crime, and the more numerate subjects made more errors than their less numerate allies. The better you are with numbers, the better you are at bending them.

The repair is small and slightly awkward: rate your confidence. Not “Citizen Kane won best picture” but “I’m a six on that.” A scale gives you somewhere to move when evidence arrives. “I was 58 percent, now I’m 46” costs nothing. “I was right and now I’m wrong” costs your self-image, so you fight it.

Luck bucket, skill bucket

Every outcome must be filed: skill, or luck? Duke’s first regular game, in a Billings basement, included a player called Nick the Greek, who reasoned that since everyone expects aces, the mathematically worst hand was the best one to play. When seven-deuce won, brilliance. When it lost, bad luck. He went broke.

We all file like Nick. Robert MacCoun found that 75 percent of drivers describing accidents blamed someone else — 91 percent in multi-car crashes, 37 percent even in single-vehicle ones. For others we flip it: Steve Bartman, one of several fans reaching for the same foul ball, collected a decade of death threats.

The counter-model is Phil Ivey, who spent the dinner after a half-million-dollar win walking another pro through every error he thought he’d made getting there. Duhigg’s rule for habit change — keep the cue and the reward, swap the routine — applies: take your ego hit for being the best credit-giver in the room.

Borrow other people, and other versions of yourself

Duke recommends a small truthseeking pod — three is enough, two to disagree and one to referee — chartered around accuracy, accountability and real disagreement. Erik Seidel put it plainly: ask me about strategy all day, but there’s no point in a poker story about something you couldn’t control. Experts betting real money on which psychology studies would replicate were right 71 percent of the time, against 58 percent for the same experts in ordinary peer review.

Then recruit yourself. Jerry Seinfeld’s Night Guy stays up late because exhaustion is Morning Guy’s problem; Suzy Welch’s 10-10-10 asks how a choice will look in ten minutes, ten months and ten years. Shown an age-progressed face in a Stanford virtual-reality study, people put $178.10 of a hypothetical $1,000 into retirement; shown their current face, $73.90. Bind your hands in advance like Ulysses: a loss limit, an automatic transfer. And run a premortem: write tomorrow’s headline announcing that you failed, then explain why. Gabriele Oettingen’s dieters who fantasized hardest about success lost twenty-four pounds less than those who pictured obstacles.

The bottom line

Grade the process, not the scoreboard: no prediction short of 0 or 100 percent is proven wrong by a single result. The present works like a chainsaw on the tree of possible futures, cutting off every branch that didn’t happen until what did happen looks inevitable — which is why writing the odds down beforehand is the only real defense. Read it if you make consequential calls under uncertainty and you’re tired of being graded on luck.