Annie Duke’s Great Lessons For Ambitious Founders

Very grateful to @AnnieDuke for her guest appearance on #StartingGreatness. She offers many great lessons for ambitious founders. Summarizing in the below thread…

1/ There are different types of “games” in business and in life. Some are based almost entirely on skill, like chess. Garry Kasparov will beat me 100 times in a row. Luck won’t rescue me. Some Games involve lots of chance and uncertainty. More like poker.

2/ If I play the world champion for an entire night, I likely lose my money. But in one hand, I might have a 40% chance of winning. Startups are more like poker than chess. They involve skill combined with uncertainty and chance across thousands of decisions, big and small.

3/ What will customers want most? Is this the right price? Will there be a recession next year? etc. Too many leaders think of decisions as all-or-nothing. They say “I assert this will happen.” Annie Duke’s books, “How to Decide” and “Thinking in Bets”, show a better way.

4/ First takeaway: Think in bets When someone asks your opinion about something where there’s high uncertainties, the correct response is “I’m not certain, but here is how I am thinking about it.”

5/ When we express our level of confidence by saying “I’m x% sure,” rather than as an all or nothing, we open the door for others to collaborate with us better, tell us what they know, and act like scientists in our shared quest for the truth.

6/ Acknowledging that decisions are bets based on our beliefs, getting comfortable with uncertainty, and redefining right and wrong are key first principles to good decision-making in uncertainty.

7/ Second Takeaway: Don’t assume a good outcome means you made a good decision; and vice versa. If you ask most people, “what were the best decisions you made in 2020?” more times than not they will answer with the things that went well. “I’m glad I bought Bitcoin in 2020.”

8/ But it’s important to avoid the mistake that Annie calls RESULTING, which is confusing the quality of an outcome with the quality of a decision. The reason is that there are way more possible futures than the one that actually happens.

9/ When you are dealing with high uncertainty, you need to consider a matrix of potential outcomes. On one dimension you have a good decision vs a bad decision. On the other dimension, you have a good outcome vs a bad outcome.

10/ Imagine I play a single hand of poker with Annie Duke and I get lucky and win. If she assumed I was better than her just because she lost a single hand, she would miss the opportunity to separate me from my money before the night is over.

11/ Or, imagine I gained false confidence after winning the first hand against Annie. I might lose all of my money even faster because it might cause me to take foolish risks.

12/ Making matters worse, we tend to let “memory creep” cause us to remember things differently from how they actually happened. We tend to think we knew it all along, especially when things go well. It seems like it was due more to our decisions and foresight than it really was.

13/ Which brings us to the final takeaway: Document your thoughts and learnings constantly In “How to Decide”, Annie tools such as a Knowledge Tracker, Decision Tree, and Premortems for every key decision.

14/ When you create a knowledge tracker, you document what you knew and when you knew it. The first thing you capture is stuff you knew or assumed before the decision. Then you focus on the new stuff you learned after making the decision and the precise time you learned.

15/ By tracking what you thought you knew and when you knew it, it helps you resist the temptation to have hindsight bias based on the outcome. It also lets you look back in time and see what factors you might have overlooked.

16/ The second tool is a Decision Tree. For every decision, you identify the reasonable set of outcomes, including the range of bad outcomes and horrible ones as well as expected outcomes and better than expected outcomes.

17/ Then, you identify your level of preference or dislike for each outcome. If one of the potential outcomes could ruin you, make sure you call that out. Next you estimate the likelihood of each outcome.

18/ You look at your options and decide which probability weighted option best matches your preferences. A decision tree also helps you understand if you framed the decision well. If an outcome is highly unexpected, maybe you can frame the possible outcomes better next time.

19/ The last important tool is a Pre-Mortem. You identify the goal you are trying to achieve or the decision you are considering. Then you figure out a time in the reasonable future. You imagine it’s the day after and things have not worked out.

20/ List up to five reasons you failed due to things within your control and list five reasons you failed due to things you can’t control. This is far better as a team exercise where everyone lists the reasons separately and then compares notes.

21/ Or, you can approach a glass half full perspective. You imagine a time in the future, but you achieved the goal. You list the five reasons you succeeded due to things in your control and five reasons you succeeded outside of your control.

22/ Startup life, like poker, is a long game, and there are going to be a lot of losses, even when you make the best possible bets. You will be happier and perform better when you realize we’ll never be sure of the future.

23/ This mindset shift changes our task from trying to be right every time to navigating through the uncertainty. By framing decision-making and leadership as a quest for truth in the face of uncertainty, you can reduce personal conflicts and counterproductive beliefs.