📖 Thinking in Bets

Making smarter decisions when you don’t have all the facts

2018. Annie Duke

Annie Duke studied psychology in college before going on to become a World Series of Poker champion. She retired from poker in 2012 (having won over $4 million in tournaments) and has spent the years since applying her poker expertise and psychology background to the art of decision making in other contexts, particularly business.

The book doesn’t necessarily introduce anything new, but rather takes existing theory and applies it to the game of Poker. And while her experience playing Poker is a fun lens through which to learn, Duke also claims it to be a very relevant one.

Popular strategist Roger L. Martin defines strategy as “making choices under uncertainty and competition.“ A definition which matches Duke’s definition of poker; “a game of decision-making under conditions of uncertainty over time”, similar again to the definition of game theory. But in Poker things are accelerated somewhat because you make hundreds of fast, consequential decisions in one session.

Resulting

Resulting is a poker term that refers to determining the quality of a decision by the quality of it’s outcome. I expanded on resulting previously, but the key point is that you can’t judge a decision by it’s outcome. A bad result does not mean it was a bad decision.

Probabilistic thinking

As the title of the book suggests, decisions are bets on the future. And as we just said, you can’t determine a decision to be right or wrong based on the outcome. And rather than viewing decisions as that black and white, 100% right to 100% wrong, she argues to “put a price” on decisions. To acknowledge the grey area by assigning a percentage confidence to any and all decisions, or even opinions and beliefs. Declaring our uncertainty in our beliefs to others makes us more credible communicators. It also invites collaboration, and opens the door for someone else to help refine our beliefs, as opposed to only leaving them with an option of telling us they think we are wrong, or right.

Biased reasoning

Our pre-existing beliefs influence the way we experience the world. They drive the way we process information. And so two people can view the same data but arrive at two different conclusions, because they processed it differently.

Motivated reasoning

Our beliefs are stubborn. Once they become lodged, they can be difficult to dislodge. They lead us to seek out evidence which confirms our beliefs, and ignore information that contradicts them. This irrational, circular information-processing pattern is called motivated reasoning.

Again, if we think of all our beliefs in absolute terms, as being either 100% right or 100% wrong, it makes it difficult to update and change our beliefs. If we are presented with information that contradicts our beliefs we either a) shift our position from 100% right to 100% wrong, or b) ignore or discredit the new information. We don’t like being wrong, it feels like an attack on our self narrative and so we go for option b and carry on.

But refusal to update our mental models like this is devastating for learning. We should be always looking to construct more accurate mental models of the world, updating and refining them with new information. We must agree that our beliefs are in progress and under construction in some way.

Thinking probabilistically helps here too. We are less likely to succumb to motivated reasoning if we are making “small adjustments in degrees of uncertainty, instead of having to grossly downgrade from ‘right’ to ‘wrong’”. “I was 80% sure on that, but now i’m probably more like 60%”.

Self-serving bias

Fielding outcomes

Outcomes are the result of skill and luck. Every time we get an outcome, we make a decision as to how much of that was attributed to the influence of our skill and decisions, and how much was attributed to the influence of luck. W

Things within our control, vs things outside of our control. And usually this is a really difficult assessment to make because there is so much missing information preventing us from knowing for sure.

Bias

But the way people field outcomes is predictably patterned: we take credit for the good stuff and blame the bad stuff on luck, what psychologists refers to as “self-serving bias.” According to Duke, this predictable fielding error, of “off-loading the losses to luck, and on-boarding the wins to skill” is probably the single most significant problem for poker players. How unlucky I was on that hand vs how I could have played that hand better. Poker players love telling their “bad beats” stories. How they were playing great, but then some card was turned over on the river and they lost. So unlucky. For Duke however, this talk wasn’t allowed in her learning group, or by her mentors. They interacted with each-other in an exploratory style, not a confirmatory style.

Pain

The thing that gets in the way of most learning is the pain. If we attribute our losing to our own poor decision making, we view this as an attack on our identity, and it’s painful. If we attribute losses to external factors out of our hands, it lets us off the hook. So the choice is to take the pain in the short term in order to help us learn for the long run. Examine our decisions, see where maybe we could have improved which would have increased the likelihood of better outcome. Or, decide that we don’t want to feel pain right now.

And in her years playing poker Duke was amazed at the extent to which people were willing to take a huge discount in the present in order to preserve that self narrative. To blame it on luck, socialise the result to the world, offload responsibility for the loss to something out there, and just stick with the strategy that lost them that hand. There are many reasons for this, namely the self-serving bias and motivated reasoning just mentioned, but another theory mentioned later in the book is “temporal discounting”, a tendency we all have to favour our present-self at the expense of our future self. A willingness to take an irrationally large discount to get a reward now, instead of waiting for a bigger reward later.

Resisting this temptation is hard work. It’s much easier to blame things on other people, than to look for ways we could have handled ourselves better.

Compounding

Duke’s rallying cry is that by being aware of these common traps, you can become a little better at decision making. And a little improvement now, can make a big difference over time. If we can recognise and catch the learning opportunities that present themselves, we can improve our decision quality, and put ourselves in a better position for the future.