By then she was married, and so she asked her husband to make some inquiries. They learned there was a bar in Billings named the Crystal Lounge where retired ranchers, construction workers, and insurance agents played poker every afternoon in the basement. It was a smoke-filled, joyless dungeon. Annie went one afternoon and loved it. She went again a few days later and walked out fifty dollars richer. “Playing poker down there was this combination of math, which I loved, and all of the cognitive science stuff I had been doing in graduate school,” Annie told me. “You could watch people try to bluff each other and hide their excitement when they got a good hand, and all these other kinds of behaviors we’d spent hours talking about in classes. Every night, I would call my brother and talk through the hands I had played that day, and he would explain my mistakes, or how someone else had figured out my game and had started using that against me or what I should do different next time.” Initially, she wasn’t very good. But she won often enough to keep going. She noticed that her stomach never hurt at the poker table.
Pretty soon, she was going to the Crystal Lounge every weekday, like a job, arriving at three P.M. and staying until midnight, taking notes and testing strategies. Her brother sent her a check for $2,400 with the agreement that he would get half her winnings. She was up $2,650 by the end of the first month even after his cut. The next spring, when he invited her to Vegas, she drove fourteen hours, bought a seat in a tournament, and by the end of the first day had $30,000 in chips.
Thirty thousand dollars was more than Annie had ever earned in a full year as a grad student. She understood poker—understood it better than many of the people she was playing against. She understood that a losing hand isn’t necessarily a loss. Rather, it’s an experiment. “The thing I had figured out by that point was the difference between intermediate and elite players,” Annie told me. “At the intermediate level, you want to know as many rules as possible. Intermediate players crave certainty. But elite players can use that craving against them, because it makes intermediate players more predictable.
“To be elite, you have to start thinking about bets as ways of asking other players questions. Are you willing to fold right now? Do you want to raise? How far can I push before you start acting impulsively? And when you get an answer, that allows you to predict the future a little bit more accurately than the other guy. Poker is about using your chips to gather information faster than everyone else.”
By the end of the tournament’s second day, Annie had $95,000 in chips. She finished in twenty-sixth place, ahead of hundreds of professionals, some with decades of experience. Three months later, she and her husband moved to Las Vegas. At some point, she called her professors at Penn and told them she wasn’t coming back.
A full minute has passed. Annie still has a pair of tens. If the FossilMan is holding a higher pair—say, two queens—and Annie stays in the hand, it’s almost certain she’ll be eliminated from the Tournament of Champions. But if she wins the hand, she’ll become the table’s chip leader.
All of the odds and probability charts floating around Annie’s head are telling her to do one thing: Match the FossilMan’s bet. But every time Annie has asked the FossilMan a question during this tournament by placing a wager, he’s answered with a highly rational response. He’s never put everything on the line without a good reason. Now, in this hand, he’s pushed all of his chips into the pot, even as Annie has raised again and again.
Annie is aware that the FossilMan knows how hard it is for her to back down at this point. He knows that, unlike some of the other people at this table, she isn’t in the Poker Hall of Fame. This is her first time in front of a million television viewers. He might even know that she’s worried she doesn’t belong here at all, that she suspects she was only invited because the TV producers wanted a woman at the table.
Annie suddenly realizes she’s been thinking about this hand wrong all along. The FossilMan has been betting as if he has a good hand because, in fact, he has a good hand. Annie has been overthinking—or, at least, she thinks she’s been overthinking. She’s not sure.
She looks at her pair of tens, looks at the $450,000 on the table, and folds. The FossilMan takes the money. Annie has no idea if she just made a good or bad choice because the FossilMan doesn’t have to show anyone his cards. Another player leans over. You completely misread the situation, he whispers to her. If you had stayed in, you would have won.
A few hands later, Annie has already folded when the FossilMan, with a ten and a nine, bets all his chips once again. It’s a smart play, the right move, but as the other cards fall on the table, they go against him. Even the smartest poker players can be undone by bad luck. Probabilities can help you forecast likelihoods, but they can’t guarantee the future. Just like that, the FossilMan is out of the tournament. As he stands to leave, he leans over to Annie.
“I know the hand you had earlier was really hard for you,” he tells her. “I want you to know that I had two kings and you made a good fold.”
When he says that, the knot of panic in her stomach melts. She’s been distracted ever since she folded against him. She’s been second-guessing herself, turning the hand over in her mind, trying to figure out if she played it right or wrong. Now her head is back in the game.
It’s normal, of course, to want to know how things will turn out. It’s scary when we realize how much rides on choices where we can’t predict the future. Will my baby be born healthy or sick? Will my fiancée and I still love each other ten years from now? Does my kid need private school or will the local public school teach her just as much? Making good decisions relies on forecasting the future, but forecasting is an imprecise, often terrifying, science because it forces us to confront how much we don’t know. The paradox of learning how to make better decisions is that it requires developing a comfort with doubt.
There are ways, however, of learning to grapple with uncertainty. There are methods for making a vague future more foreseeable by calculating, with some precision, what you do and don’t know.
Annie is still alive at the Tournament of Champions. She has enough chips to stay in the game. The dealer gives each player their next cards and another hand begins.
II.
In 2011, the federal Office of the Director of National Intelligence approached a handful of universities with grant money and asked them to participate in a project “to dramatically enhance the accuracy, precision, and timeliness of intelligence forecasts.” The idea was that each school would recruit a team of foreign affairs experts, and then ask them to make predictions about the future. Researchers would study who made the most accurate forecasts and, crucially, how they did it. Those insights, the government hoped, would help CIA analysts become better at their jobs.
Most of the universities that participated in the program took a standard approach. They sought out professors, graduate students, international policy researchers, and other specialists. Then they gave them questions no one yet knew the answers to—Will North Korea reenter arms talks by the end of the year? Will the Civic Platform party win the most seats in the Polish parliamentary elections?—and watched how they went about answering. Studying various approaches, everyone figured, would provide the CIA with some fresh ideas.
Two of the universities, however, took a different tack. A group of psychologists, statisticians, and political scientists from the University of Pennsylvania and the University of California–Berkeley, working together, decided to use the government’s money as an opportunity to see if they could train regular people to become better forecasters. This group called themselves “the Good Judgment Project,” and rather than recruit specialists, the GJP solicited thousands of people—lawyers, housewives, master’s students, voracious newspaper readers, and enrolled them in online forecasting classes that taught them different ways of thinking about the future. Then, after the training, those participants were asked to answer the same foreign affairs questions as the experts.