by Annie Duke
Look how quickly you can begin to redefine what it means to be wrong. Once we start thinking like this, it becomes easier to resist the temptation to make snap judgments after results or say things like “I knew it” or “I should have known.” Better decision-making and more self-compassion follow.
The public-at-large is often guilty of making black-and-white judgments about the “success” or “failure” of probabilistic thinking. When the UK voted to leave the European Union (“Brexit”) in July 2016, it was an unlikely result. Betting shops had set odds heavily favoring a vote for Remain. That does not mean the betting shops had an opinion that Remain would win the day. The goal of the bookmaker is to make sure the amount of money bet on either side is equal, so that the losers essentially pay the winners while the bookmaker just takes their fee. They aim to have no stake in the outcome and adjust the odds accordingly. The bookmaker’s odds reflect the market’s view, essentially our collective best guess of what is fair.
That didn’t stop even sophisticated people from resulting, declaring after the vote came in Leave that the bookmakers made a mistake. The chief strategist at one Swiss bank told the Wall Street Journal, “I can’t remember any time when the bookies were so wrong.” One of America’s most famous lawyers and professors, Alan Dershowitz, made this same error. Asserting in September 2016 that the Clinton-Trump election was too difficult to make any predictions about, he said, “Think about the vote on Brexit. Virtually all the polls—including exit polls that asked voters how they voted—got it wrong. The financial markets got it wrong. The bookies got it wrong.”
Just like my spectator, Dershowitz missed the point. Any prediction that is not 0% or 100% can’t be wrong solely because the most likely future doesn’t unfold. When the 24% result happened at the final table of the charity tournament, that didn’t reflect inaccuracy about the probabilities as determined before that single outcome. Long shots hit some of the time. Blaming the oddsmakers or the odds themselves assumes that once something happens, it was bound to have happened and anyone who didn’t see it coming was wrong.
The same thing happened after Donald Trump won the presidency. There was a huge outcry about the polls being wrong. Nate Silver, the founder of FiveThirtyEight.com, drew a lot of that criticism. But he never said Clinton was a sure thing. Based on his aggregation and weighting of polling data, he had Trump between 30% and 40% to win (approximately between two-to-one and three-to-two against) in the week before the election. An event predicted to happen 30% to 40% of the time will happen a lot.
Being a poker player, I’ve played out more two-to-one shots in my tournament career than I could possibly count. A lot of those have been situations where the tournament was on the line for me. If I lose the hand, I’m out of the tournament. If I win, I earn a huge pot, maybe even winning the entire tournament. I know viscerally how likely 60–40 and 70–30 favorites are to lose (and, of course, the opposite). When people complained that Nate Silver did his job poorly because he had Clinton favored, I thought, “Those people haven’t gotten all their chips in a pot with a pair against a straight draw and lost.” Or, more likely, they’ve had those things happen throughout their lives and didn’t realize that’s what 30% or 40% feels like.
Decisions are bets on the future, and they aren’t “right” or “wrong” based on whether they turn out well on any particular iteration. An unwanted result doesn’t make our decision wrong if we thought about the alternatives and probabilities in advance and allocated our resources accordingly, as my client the CEO and Pete Carroll both did. It would be absurd for me, after making a big bet on the best possible starting hand (a pair of aces) and losing, to spend a lot of time thinking that I was wrong to make the decision to play the hand in the first place. That would be resulting.
When we think probabilistically, we are less likely to use adverse results alone as proof that we made a decision error, because we recognize the possibility that the decision might have been good but luck and/or incomplete information (and a sample size of one) intervened.
Maybe we made the best decision from a set of unappealing choices, none of which were likely to turn out well.
Maybe we committed our resources on a long shot because the payout more than compensated for the risk, but the long shot didn’t come in this time.
Maybe we made the best choice based on the available information, but decisive information was hidden and we could not have known about it.
Maybe we chose a path with a very high likelihood of success and got unlucky.
Maybe there were other choices that might have been better and the one we made wasn’t wrong or right but somewhere in between. The second-best choice isn’t wrong. By definition, it is more right (or less wrong) than the third-best or fourth-best choice. It is like the scale at the doctor’s office: there are a lot more choices other than the extremes of obesity or anorexia. For most of our decisions, there will be a lot of space between unequivocal “right” and “wrong.”
When we move away from a world where there are only two opposing and discrete boxes that decisions can be put in—right or wrong—we start living in the continuum between the extremes. Making better decisions stops being about wrong or right but about calibrating among all the shades of grey.
Redefining wrong is easiest in situations where we know the mathematical facts in advance. In the charity-tournament final-table example with the players’ cards faceup, or when I get all my chips in with the best possible starting hand, the hidden information is removed. We can make a clear calculation. If we have that unquestionably right and make an allocation of resources (a bet) on the calculation, we can more naturally get to “I wasn’t wrong just because it didn’t turn out well and I shouldn’t change my behavior.” When the chances are known, we are tethered more tightly to a rational interpretation of the influence of luck. It feels a little more like chess that way.
There is no doubt it is harder to get there when we add in hidden information on top of the influence of luck. Untethered from seeing what the coin actually looks like, we are more likely to anchor ourselves to the way things turned out as the sole signal for whether we were right or wrong. We are more likely to declare, “I told you so!” or “I should have known!” When we start doing that, compassion goes out the window. Just ask Pete Carroll.
Redefining wrong allows us to let go of all the anguish that comes from getting a bad result. But it also means we must redefine “right.” If we aren’t wrong just because things didn’t work out, then we aren’t right just because things turned out well. Do we win emotionally to making that mindset trade-off?
Being right feels really good. “I was right,” “I knew it,” “I told you so”—those are all things that we say, and they all feel very good to us. Should we be willing to give up the good feeling of “right” to get rid of the anguish of “wrong”? Yes.
First, the world is a pretty random place. The influence of luck makes it impossible to predict exactly how things will turn out, and all the hidden information makes it even worse. If we don’t change our mindset, we’re going to have to deal with being wrong a lot. It’s built into the equation.
Poker teaches that lesson. A great poker player who has a good-size advantage over the other players at the table, making significantly better strategic decisions, will still be losing over 40% of the time at the end of eight hours of play. That’s a whole lot of wrong. And it’s not just confined to poker.
The most successful investors in start-up companies have a majority of bad results. If you applied to NASA’s astronaut program or the NBC page program, both of which have drawn thousands of applicants for a handful of positions, things will go your way a minority of the time, but you didn’t necessarily do anything wrong. Don’t fall in love or even date anybody if you want only positive results. The world is structured to give us lots of opportunities to feel bad about being wrong if we want to measure ourselves by outc
omes. Don’t fall for it!
Second, being wrong hurts us more than being right feels good. We know from Daniel Kahneman and Amos Tversky’s work on loss aversion, part of prospect theory (which won Kahneman the Nobel Prize in Economics in 2002), that losses in general feel about two times as bad as wins feel good. So winning $100 at blackjack feels as good to us as losing $50 feels bad to us. Because being right feels like winning and being wrong feels like losing, that means we need two favorable results for every one unfavorable result just to break even emotionally. Why not live a smoother existence, without the swings, especially when the losses affect us more intensely than the wins?
• • •
Are you ready to really wrap your arms around uncertainty, like great decision-makers do? Are you ready to embrace this redefinition of wrong, and to recognize you are always guessing and that those guesses drive how you place your resources? Getting comfortable with this realignment, and all the good things that follow, starts with recognizing that you’ve been betting all along.
CHAPTER 2
Wanna Bet?
Thirty days in Des Moines
During the 1990s, John Hennigan, an eccentric gambler who had been making a living by his wits and skills in poker and pool for several years, moved from Philadelphia to Las Vegas. His reputation and nickname, “Johnny World,” preceded him, due to his already exceptional skills and willingness to bet on anything. His talent has stood the test of time: he is a legendarily successful player in high-stakes games, and in major poker tournaments has earned four World Series of Poker bracelets, a World Poker Tour championship, and more than $6.5 million in prize money.
John was a perfect match for Las Vegas. He arrived already in rhythm with the town: sleeping all day and spending all night in poker games, pool halls, bars, and restaurants with adventurous, like-minded peers. He quickly found a group of professional gamblers with similar interests, many from the East Coast.
Although John and Vegas seemed made for each other, he had a love-hate relationship with the lifestyle. Playing poker for a living has the allure of giving you the freedom to make your own schedule but, once it boils down to your per-hour net advantage, you are tethered to putting in the hours. You’re “free” to play or not play whenever you want, but you can feel compelled to punch a clock. Worse, the best games are at night, so you’re working the graveyard shift. You get out of rhythm with the rest of the world, never see the sun, and your workplace is a smoke-filled room where you can’t even see outside. John felt this keenly.
One night, John was in a high-stakes poker game and the talk between hands somehow included the state capitol of Iowa, Des Moines. John had never been there or seen much of the Midwest, so he was curious about what life in Des Moines might be like—a “normal” life that increasingly seemed foreign to him, waking up in the morning and living in the daylight hours. This led to some good-natured ribbing as the other players in the game imagined the prospect of a nocturnal action junkie like John in a place that seemed, to them at least, like the opposite of Las Vegas: “There’s no gambling action.” “The bars close early.” “You’d hate it there.” Over the course of the evening, the discussion shifted to whether Hennigan could even live in such an unfamiliar place.
As is often the case with poker players, a conversation about a hypothetical turned into an opportunity to propose a wager. What would the stakes have to be for Hennigan to get up from the table, catch a flight, and relocate to Des Moines? If he took such a bet, how long would he have to live there?
John and the others landed on a month in Des Moines—a real commitment but not a permanent exile. When he seemed willing to, literally, walk out of a poker game and move 1,500 miles to a place he had never been, the other players added a diabolical condition to the negotiation: he would have to confine himself to one street in Des Moines, a street with a single hotel, restaurant, and bar, where everything closed at 10 p.m. That enforced idleness would be a challenge for anyone, regardless of the location. But for someone like John, a young, single, high-stakes gambler, this might actually count as torture. John said he would take such a challenge if they made one concession: he could practice and play at a nearby golf course.
After agreeing on the conditions, they still had to negotiate the size of the bet. The other players needed a number that was large enough to entice John to accept the wager, but not so large that it would entice John to stay even if he really hated it in Iowa. As one of the most successful cash-game players in Las Vegas, a month in Des Moines could cost John, potentially, six figures. On the other hand, if they offered him too large of an upside to stay in Des Moines, he would certainly endure the discomfort and boredom.
They settled on $30,000.
John considered two distinct and mutually exclusive alternatives: taking the bet or not taking the bet. Each came with new risks and new reward potentials. He could win or lose $30,000 if he took the bet (or win or lose greater dollar amounts at the poker table if he turned it down). He could also win to the decision to move to Des Moines long after the bet was over, if he used the golf-practice time to improve his chances gambling at high-stakes golf. He could further his reputation of being willing to bet on anything and being capable of anything, a profitable asset for professional gamblers. He also had to think about the other, less quantifiable things he might value. How much might he like the pace of life? How would he value taking a break from the action? Would he become more relaxed experiencing the more traditional schedule? Was the break worth it to take the big pay cut from not being able to play poker for a month? And then there were the real unknowns. He might just meet the love of his life on that one street in Iowa. He had to weigh all of this against the opportunity costs of leaving Vegas—money from lost earning opportunities, nights missing doing the things he enjoyed, and even perhaps missing meeting the love of his life at the Mirage during that month.
Johnny World moved to Des Moines.
Was a month of detox away from the nightly life of a high-stakes Vegas pro going to be a blessing or a curse?
It took just two days for him to realize that it was a curse. From his hotel room in Des Moines, John called one of his friends on the other side of the bet and tried to negotiate a settlement. Just as parties in commercial lawsuits often settle before trial, in the gambling world negotiated settlements are common. What was particularly funny about John’s call was that his opening offer was that the others pay him $15,000 to spare them the cost and indignity of losing the whole amount. He argued that since he was already in Des Moines, he was clearly capable of waiting out the month to get the full amount.
The other bettors, literally, were not buying it. After all, John made this offer after only two days. That was a pretty strong signal that not only would they likely win the bet, but they might earn a return (in fun) by needling John while he served out his sentence.
Within a few days, John agreed to pay $15,000 to get out of the bet and return to Vegas. John proved, in spectacular fashion, that the grass is always greener.
We’ve all been to Des Moines
The punch line of the John Hennigan–Des Moines story—“after two days, he begged to get out of it”—made it part of gambling folklore. That punch line, however, obscures how usual the underlying analysis about whether to move was. The only real difference between Johnny World’s decision to move to Des Moines and anyone else’s decision to relocate or take a job was that he and the poker players made explicit that the decision was a bet on what would most improve their quality of life (financial, emotional, and otherwise).
John considered two distinct and mutually exclusive alternative futures: taking the bet and living for a month in Des Moines, or not taking the bet and staying in Las Vegas. Any of us thinking about relocating for a new job has this same choice between moving, with the potential to earn the money being offered, or staying where we are and maintaining the status quo. How does the new job pay co
mpared to what we have now? There are plenty of things we value in addition to money; we might be willing to make less money to move to a place we imagine we would like a lot better. Will the new job have better opportunities for advancement and future gains, independent of short-term gains in compensation? What are the differences in pay, benefits, security, work environment, and the kind of work we’d be doing? What are we giving up by leaving our city, colleagues, and friends for a new place?
We have to inventory the potential upside and downside of taking the bet just like Hennigan did. That his $30,000 wasn’t a sure thing doesn’t make his decision distinct from other job or relocation decisions. People take jobs all the time where a large portion of the compensation is contingent. In many businesses, compensation includes bonuses, stock options, or performance-based pay. Even though most people don’t have to consider losing $30,000 when they take a job, every decision has risks, regardless of whether we acknowledge them. Even a set salary is still not “guaranteed.” We could get laid off or hate the job and quit (as John Hennigan did), or the company could go out of business. When we take a job, especially one promising big financial rewards, the commitment to work can cost us time with our family and affect those relationships, a costly if not losing compromise.
In addition, whenever we choose an alternative (whether it is taking a new job or moving to Des Moines for a month), we are automatically rejecting every other possible choice. All those rejected alternatives are paths to possible futures where things could be better or worse than the path we chose. There is potential opportunity cost in any choice we forgo.
Likewise, the players on the other side of that bet, risking $30,000 to see if John would live a month in Des Moines, thought about similar factors that employers consider in making job offers or spending money to create enticing workplace environments. The poker players had to strike a fine balance in offering that bet to Hennigan: the proposition had to be good enough to entice him to take the bet but not so good that it would be guaranteed to cost them the $30,000.