7 Questions with Poker Legend Annie Duke on Decision-Making

by The Campaign Workshop

Poker Legend, Annie Duke on Decision-Making

Annie Duke shares her insight on making smarter decisions in life and in politics

Annie Duke is a decision-making expert and a former professional poker player. She won her first World Series of Poker, as well as the World Series of Poker Tournament of Champions, in 2004. In 2010, she went on to win the NBC National Heads-Up Poker Championship. Annie Duke founded a nonprofit called How I Decide in 2014 that helps young people develop stronger decision-making skills. Prior to her professional poker playing career, Annie Duke was awarded a National Science Foundation Fellowship and studied cognitive psychology at the University of Pennsylvania. She recently authored Thinking in Bets: Making Smarter Decisions. We talked with Annie Duke about how we make decisions, and how we can make better ones in life and in politics.

1. How much should people trust their intuitions in making decisions?

Be wary of any justification of a decision that goes something like, “My gut is telling me this is right,” because intuition as a decision-making tool is only as good as our ability to communicate to another human being why our gut is telling us something.

Intuition is an automatic, reflexive process by which we come to some understanding without the aid of conscious reasoning. Our gut is certainly an indispensable tool in our decision-making arsenal. We make so many decisions in any given day that, if we stopped and deliberated about every decision, we’d be paralyzed. We would never make it out of the door in the morning. Intuition is necessary simply because we don’t have the luxury of enough time to fully deliberate on every decision.

But our gut is also deeply subject to bias. If we never make intuition accountable to conscious reasoning, we’re never going to notice it. Unchecked, intuition will lead us astray.

If we can't adequately explain the reason that we think the intuitive path is the best path, then we should question our intuition. If we can’t explain a decision such that another person could learn from it and repeat that decision for themselves, we should reexamine the intuitive choice. And “because my gut is telling me so” should never count as an adequate explanation.

By making our gut accountable to conscious reasoning, the intuitive choice will more often converge with the rational choice. We will spot more of the times when intuition leads us astray and reinforce our gut feeling when it leads us to a good decision. This refines and hones our gut, sharpening our intuition.

2. How can folks combat their cognitive biases to become more effective decision makers?

This is a “good news-bad news” situation. The bad news is that we know several things that don’t help much in combating bias. First, knowing about such biases doesn’t help much. We have big blind spots to our own biases. Second, being smart doesn’t help. In fact, in many instances, extra expertise or brain power just means more mental agility to spin a biased story that fits our preferred narrative.

The good news is that we are good at spotting biases in other people. You can see this on both sides of the aisle in politics. One side will claim bias on the other, often correctly. The other side will call the first side out on bias, also often correctly. But nobody seems to be recognizing the biases in their own thinking.

We can use that to our advantage to recruit other people to help spot our biases and we can do the same for them. The group can offer each other alternate perspectives, fill holes in the information we have, offer up alternative hypotheses, and argue the opposing side.

A decision pod does two great things for you. First, if the group is committed to honestly exploring all sides of an issue – rather than being an echo chamber to confirm each other’s beliefs – the decisions you work through together are going to be better. Second, the mere existence of a group that we know, in advance, will hold us accountable to our beliefs and decisions helps us to be more rational in our thinking on our own. If we know we’ll have to answer to the group later for the decisions we make, the group gets in our head in a good way, making in-the-moment gaffes less likely and sharpening our intuitive responses.

3. How does embracing uncertainty and thinking in bets allow people to make better decisions? How does this play out in politics?

Thinking in bets is, literally, the process of imagining if we’d be willing to bet on our beliefs and predictions. Thinking about whether (and how much) we’d be willing to bet triggers us to vet our beliefs. To win at a bet, we do better if we consider what information might confirm or disconfirm our belief, think about why a belief might not be true, and imagine alternative hypotheses.

Thinking in bets acts as an antidote to our tendency to entrench and polarize. It causes us to recognize where and why we might be uncertain, leading us to calibrate and refine our beliefs. And the more accurate the beliefs we hold, the better our decisions will be because every decision we make is informed by our beliefs.

Imagine how thinking in bets can inoculate us against fake news or disinformation. The danger of fake news isn’t that it will change our minds and convince us of things that don’t gel with our world view. The danger is that it will cause us to retreat further into the beliefs we already hold, making those views more and more extreme. Fake news silos us.

What if we had to bet on whether a story we read was true? We’d start the vetting process: being more critical in evaluating sources and looking at the case made by people potentially betting against us. That open mindedness would serve us well in politics right now.

A second example of how thinking in bets could help political discourse is in our attitudes about polls. Polls are predictions. They are probabilistic. They are not black-and-white, right-and-wrong. Unfortunately, the public looks to polls for answers that are certain (and pundits tend to oblige).

We can see, from the last presidential election, how looking to polls for certain answers mucks things up. Nate Silver had Trump at a 30-40% chance of winning in the week leading up to the election. When Trump won, a lot of people declared, “Nate Silver got it wrong.” They refused to recognize that predictions about the future are, by definition, uncertain. Do you know how often 35% happens? That’s how often a major league batting champion gets a hit. That’s how often in a week Monday and Tuesday and half of Wednesday come around.

Imposing false certainty on a series of predictions has made it easier for people to dismiss polls, calling them fake news. But if we had all been better at thinking in bets, we would have recognized the uncertainty in the outcome of the election in the first place and no one would be declaring the polls “wrong.”

4. Can political campaigns harness lessons from behavioral/decision sciences to better persuade voters?

Our concept of “identity” is a driving force in our decisions and our worldview. We need to resolve conflicts in a way that protects our fundamental view of ourselves. Because we are tribal in nature, our political party becomes our tribe and, consequently, part of our identity. To encourage party members to vote, a sound behavioral strategy is to pose their voting decision as a matter of identity, rather than as a matter of action. On the Democratic side, for example, it would be more powerful to ask, “Are you a Clinton voter?”, than “Are you planning to vote for Clinton?” In the first case, you’re creating an identity and asking if they will behave in a way consistent with that identity. In the second, you’re asking about an action they may or may not take.

You can also encourage voters to make the act of voting part of their identity. (That’s where the ubiquitous “I voted today” stickers came from.) That’s not strictly a party-identification act, but it’s likely to be party-favorable. We already know that people are reluctant to cross their party. In attempting to persuade party members who don’t like the candidate, an alternative could be to appeal to an identity that supersedes their political party (“Americans” or “citizen-participants in government”).

(I highly recommend the work of Todd Rogers, a professor of public policy at the Harvard Kennedy School, on how you can harness the ideas of behavioral economics to persuade voters.)

5. How can voters make smarter decisions?

One example of how thinking probabilistically about voting can lead to smarter decisions involves crossover voting in primary elections. If you’re voting in a primary election and your choice within your preferred party is inconsequential (your preferred candidate has a huge lead, or the candidates are close and you are indifferent to which becomes the party’s nominee), consider crossover voting. By voting in the opposing party’s primary, you can vote for the candidate that is the most palatable to you in case the opposing party wins the general election.

I live in Pennsylvania, a battleground state, and I wrote a blog about maximizing the value of your vote by crossover voting before the Pennsylvania primary in 2016. A friend of mine told me he followed that advice. He was a lifelong Democrat and his preferred candidate was Hillary Clinton, who was leading in the polls by fifteen percent. Since his preferred candidate was ahead by so much, he felt his vote didn’t matter so much on the Democratic side. 

There were a couple Republican candidates he was very uncomfortable with, but one he preferred if the next president was a Republican. He changed his registration to Republican and voted in the primary for John Kasich.

Few people do this, and what happened to my friend later explains why. When he told his mother that he changed his registration – just for the primary – to vote for the most palatable Republican candidate, she was apoplectic.

“ARE YOU CRAZY? If you get hit by a bus tomorrow, that means you’ll die a Republican!”

As long as we place our party – our tribe – above our beliefs, we’re in danger of missing ways to make our vote more valuable.

6. Why do people think they can predict outcomes with 100 percent certainty?

People don’t necessarily think they can predict outcomes with 100% certainty. Deep down, we all recognize that the future is uncertain, but that makes us uncomfortable. We associate certainty with knowledge and confidence, and we feel more comfortable thinking things happen for reasons (and not randomly) and that we are in control of how our lives turn out. We were built to think in cause and effect. When our ancestors were on the savannah and heard rustling, it could be the wind or it could be a lion. Our survival depended on making judgments about those connections – and if we weren’t sure, it was safer to bet on lion.

Because we avoid thinking about uncertainty, we default to thinking about beliefs and predictions as 0% or 100%, black and white, right and wrong. That’s what happened in the interpretation of the 2016 presidential polls. That’s why pundits don’t waffle. They fulfill our need to think that the world is a certain and predictable place.

7. You run a nonprofit that helps young people hone their critical thinking and decision-making skills. How do you go about teaching these skills to kids? Is it harder to get adults to recalibrate?

How I Decide provides programming to underserved middle schoolers to improve each stage of the decision-making process, from emotional control to probabilistic thinking.

Our Mindful Choices program improves decision fitness. If someone is in an emotional state, it doesn't matter how much they know about decision-making; they won’t be able to execute on it. Mindful practice improves the ability to calm down in the moment of a decision and get into a more rational place.

HabitWise teaches kids about habit formation. Because we make so many decisions automatically (intuition!), kids need the ability to identify good habits from bad habits. Once they've identified the habits that they want to change, HabitWise helps them develop the tools to change those habits.

In general, all of these areas are places where adults need improvement as well. Adults can learn to be better decision-makers. But, just like kids, they need help. The fastest way for adults to improve is to form a good decision group committed to open-minded discourse, representing diverse viewpoints, and holding each member accountable for biased thinking.


8. Do you play cards with your kids? What do you play?

Yes, as long as you count Apples-to-Apples.

9. You were the runner-up on Celebrity Apprentice in 2009. Any thoughts on what kind of strategy you would use to play poker against President Trump?

I’m pleading the fifth on this one. I hope someday I may actually play against him and I don’t want to reveal my strategy.

10. Have you read anything lately that you think sheds light on the American electorate?

I recently read a great paper by Jay Van Bavel and Andrea Pereira from NYU that delves into how political parties are tribal in nature. They identify the social benefits of political party affiliation: “For example, political rallies and events can satisfy belonging needs; party elites, partisan media, and think tanks provide policy information; in-group members model norms for action; electoral success confers status and power; and party policy provides guidance on the appropriateness of values.” Our political affiliation has been shown to affect how we do math equations, our memory, and even how we perceive the world.  

I think the most surprising (and counter-intuitive!) finding is that when our ideology conflicts with a policy of our political party, we resolve the conflict by changing our belief rather than changing our political party.

Political party, not policy, is in the driver’s seat when it comes to our identity.