Skip to main content

E226 | The Science Behind Forecasting And Making Decisions with Annie Duke

Hard decisions are part of every aspect of human life. In business, these shape a company’s future and define its success if done right. We often praise the hiring process in a company, as it constitutes a decision based on precise forecasting and analysis. However, we often brush off the equally important decision of letting go of someone, as we generally think it’s done in one way; subjective decision, end-of-contract meeting, and emptying a desk. We don’t realise that letting go of an employee should be subject to the same amount of well-thought analysis, as it is as strategic for the company as hiring. 

This week on the Melting Pot, we talked with and learned from Annie Duke, an ex-professional poker player and author of two books, Thinking in bets and Quit. The first makes a parallel between poker and business, and covers ways in which we could bring the critical decision-making process from gambling into the entrepreneurial adventure. The latter is a gem that helps us know when to call it quits. More specifically, it helps decision-makers discover the neuroscience behind firing people and how to do it right. She proposes a precise methodology to know where an employee is not a good fit anymore and how to let them understand that it is time for them to quit. 

After earning the title of ‘The Duchess of Poker’, she now focuses on cognitive-behavioural decision authorship and coaches businesses in making the right decisions in their environment. 

Download and listen to this fascinating episode where Annie shares the journey that led her to coaching decision-makers, interesting concepts such as loss aversion, and what aspects of the human cognitive bias can affect our forecasting. 

In today’s episode: 

  • The way our mind works when it comes to quitting things 
  • The loss aversion bias in decision making
  • Specificity vs Sensitivity in decision making 
  • Thinking in Bets, and playing poker
  • The role of luck in success


The science behind forecasting liabilities and the art of letting them go with Annie Duke, author of “Quit”. 

Annie Duke is an author, speaker, and consultant in the decision-making space. She is an ex-poker player and lapsed academic. During Covid, she was invited to research forecasting, as she was already familiar with this space. Hoping to create good training to help novices improve, she produced a methodology to become a super forecaster by providing an environment that will push them towards becoming better at it. She did four large-scale studies that generated great results and was granted permission to use them for her academic dissertation. Annie is now enrolled in her Ph.D. and awaiting to defend her dissertation. 

“There are two different types of forecasting: one is prospective, meaning how we think things will unfold over the coming months (e.g. thinking about what route I want to take for work). These forecasts are always involved in our decision-making; then, what’s called counterfactual forecasting, which is retrospective. There is the classic one, what if we had killed baby Hitler? How would that have affected the way things would have unfolded? It’s hypothetical, but you could know the answer to some things.”

A lot of how we think about making decisions is actually based on counterfactual forecasting. We ask the “what-if” questions in business which are pretty common. Like the “what if Covid didn’t happen? What would that mean for my business?”. This is no easy task because we have lots of cognitive biases. Annie Duke started this training to help people get better at that and overcome their cognitive bias by bringing a rational scientific aspect to forecasting. 

It’s all about working out what to do going forward by thinking about the past and imagining different outcomes. We could apply the same thing in forecasting geopolitical outcomes for the Russia-Ukraine war. We could define the distribution of results based on probabilities, where we define the least probable situation and the most probable one based on facts. 

The way our mind works when it comes to quitting things 

Annie recognises that there are many forces at work when deciding to quit things. First is the sunk cost fallacy, where we consider what we’ve spent and decide whether we want to spend more. And then there is the identity that we usually tie with our actions. In this case, stopping and moving on to something else becomes even harder, particularly when going against the tide. 

“The interesting thing is that when the owner of the decision goes away, it becomes much easier to quit. So if Putin were to be deposed, the new people who came in could say: ‘that wasn’t our decision. This was stupid, and it’s time for everybody to go home”. 

Annie highlights the important role that identity plays in making decisions. When our identity is highly correlated with what we do, we tend to make decisions based on what defines us, not what is right in that situation – even if it’s wrong. In business, this happens when an employee links his identity with his role within the company. If the company grows, or the environment changes, and they need to embody a different role, they struggle even when it might be an improvement for them. 

The loss aversion bias in decision making 

Loss aversion is part of the prospect theory developed by Amos Tversky and Daniel Kahneman. When we want to start something new, we focus on the downside more than the upside. The equivalent loss feels twice as bad as the gain, emphasising the loss and preventing us from moving forward. The result is an asymmetric evaluation of choices, generally making the most comfortable choice because starting something new has a stronger downside than its upside. That in itself constitutes a cognitive bias. 

“It’s a very weird quirk of our cognition, but it’s a huge problem. I started coaching people into that by asking them if it would be better if they had nobody in the role. I try to get them away from “what if you hire someone new and they’re bad” to “what if you had no one in the role”. What happens when you have someone who’s not performing in the role is that the answer to that question is always yes. This helps them get out of that world aversion and gets them to focus on subtraction instead of addition because addition is what gets their mind starting.” 

Annie shares that it’s better to have no one than a blocker in a role. Someone who will make your business struggle instead of run smoother. We are naturally bad at letting people go, so it’s important to learn how to do it properly, especially in a fast-growing business. The vital reaction to a situation like this is to set deadlines or, in Annie’s words: “Kill criteria”. 

Defining a timeframe in which the decision maker is okay with the bad employee situation is crucial. Annie defines with her clients what signals would tell if that person is not a good fit for the company and which ones would tell you if that person could turn it around. These become inputs necessary to get on the good side of the equation. 

This generates a collaborative way in which the decision-maker provides the support needed for the employee to turn it around and makes them aware that the situation is crucial. 

“The people I work with don’t do that because I warn them in advance. I say, you’re going to have this conversation, and they won’t do it during the conversation, but maybe a day or two later, probably about 60% of the time, they will resign. And you’re going to let them.” 

“What happens is that the employee is going to realise that they are never going to achieve the good stuff, and they leave. It’s better for everybody and, most importantly, for the employee. You’re denying them the feedback they need for their career improvement and fulfilment”.

Specificity versus sensitivity in decision making

People’s perception of risk is not so good, and it is mainly influenced by past events. Specificity is getting something exactly right. It might create a lot of false negatives, as opposed to sensitivity, which is overreacting to the environment to evade any risk. In life, as in business, we want a good balance. Not getting too specific that it gets us into traps, and not getting too sensitive that we are scared of everything around us. 

“We can think about this in terms of COVID tests. Specificity is going to be like it’s going to get if you have COVID, but it will have a very high false negative rate if you’re overly specific. In contrast, sensitivity wants to be highly sensitive to COVID, but that will give you lots of false positives. And what we’re trying to do is come up with a test that’s both specific and at the appropriate level of sensitivity.”

This aspect of our human nature makes our decision-making hard in most cases. Especially when it comes to forecasting, when coupled with loss aversion, it can lead to a much biased cognitive experience and, subsequently, a bad decision.

Annie proposes a solution centred around data. The probability of that forecast happening, along with the percentage of times that it happened, should give any decision-maker a good evaluation of risk in any given situation. It’s critical to look at the base rates and how it evolves through the years. 

Thinking in Bets and playing poker 

Annie started playing poker because she had left school and needed money. She developed a golden hand at it and became a champion. That experience taught her a lot about decision-making and risk-taking, which she translated into a book. She always saw poker more as trading than gambling. And what’s traded is the value of one’s hand. 

The role of luck in success 

According to Annie, some successful people make a fortune based on one lucky decision to invest early in a project, but that doesn’t necessarily make them geniuses. She highlights the importance of bringing a process into making that decision, rather than purely speculating on a company or investing a lot of money and then miraculously getting excellent results. 

Annie recognises that luck sometimes plays an immense role in decision-making, and that shouldn’t make us arrogant. Instead, one should implement a strategy to further grow that investment with a process. 

Laying process on top of luck can add great accuracy to business decision-making.

“People can hide in luck, which makes it that so that they don’t necessarily have to examine the quality of their decision.” 

The thing that Annie wished she had known before 

Annie admits that she wished she had known that her advisors were okay with her quitting graduate school, as she carried a lot of guilt and shame around that.

“I felt that I had failed the program. I felt a lot of shame around that. And then, when I reconnected with my advisor, it was wonderful. She made it very clear that she was just happy for me. And so I’ve been walking around feeling like I had really let this woman down, who I loved, and feeling a lot of shame about it, guilt about it.”

Book recommendations 

“Super Forecasting” by Phillip E.Tetlock and Dan Gardner 

“Algorithms to Live By” by Brian Christian and Tom Griffiths 

“The Modern Thinker” by Alex Sangha 

“ The Success Equation: Untangling Skill and Luck in Business, Sports, and Investing” by Michael J. Mauboussin 

“Thinking Fast and Slow” by Daniel Kahneman

“Nudge” by Richard H. Thaler

“How to change” by Katy Milkman

Enjoyed the show? Leave Us a Review

    Fantastic! Give us your details and we'll call you back

      Enquiry | Scaling Up Master Business Course