Why We Screw Up

And What We Can Do About It

A mistake that wasn’t

During the last live game I played prior to COVID, I was situated at an aggressive table at a local casino. The game in question was No Limit Texas Hold’em. The player to my left (“Villain”), who had built up his stack from $1k to around $3k over the course of several hours, had just lost nearly half of it in a huge pot against the drunken player to my right (“Drunk”). 

Despite having played reasonably throughout the night, Villain was clearly pissed at losing a big hand to Drunk, who to this point had downed five or six Jack and Cokes and was showing no signs of slowing down. After the dealer confirmed the winning and losing hands, Villain threw his former chips angrily into the pot and began mumbling incoherently to himself. 

I’d seen these signs before. While it was too early to know for certain, Villain was showing the classic symptoms of going on tilt. Tilt is a poker player’s most worrisome affliction, causing erratic behavior and an unrelenting desire to part with your money. However, unlike the Wuhan Coronavirus, tilt’s incubation period can be as short as the next hand – minutes or less. If my diagnosis was accurate, I knew it was only a matter of time before Villain gave away his remaining stack to someone at the table. I just had to hope that it’d come my way.

Several uneventful hands later, with Villain still mumbling to himself,  I looked down at my two hole cards and squeezed: Ace……..Ace. Pocket aces, the best starting hand in Hold’em. This was my opportunity. There were a couple of callers, an early raise to $50, and a re-raise by Drunk to $250. It was on me.

Normally the decision here is easy: re-raise with my Aces and get Drunk (who had just ordered another Jack and Coke) heads-up. But there were three factors that suggested I deviate from that strategy: 1) Drunk had been raising aggressively, so it’s unlikely he had a strong hand that would put in more money 2) Drunk had less money than Villain, meaning I would win more if Villain decided to play back at us 3) Given how the Drunk was playing, I thought it likely that Villain would put in a big raise against him even if I also called (called a “squeeze” play), and 4) Based on Villain’s physical mannerisms I believed he had a fairly strong hand, making it even more likely that he’d attempt a squeeze. Given those factors, I believed my best play was to call and hope that Villain went all-in. 

The hand unfolded exactly as I had hoped. I called Drunk’s $250, Villain again mumbled something to himself then shoved his $1.7k stack into the pot. Everyone else (including Drunk) folded around to me, and I quickly called and tabled my Aces. Villain looked at my cards, grimaced, and shook his head. “I need a 9” he said, meaning he had two 9’s in his hand and that only one of the two remaining nines in the deck would give him the win. The dealer dealt the flop and the first card in the window was…………a 9. Crap. Two low cards rounded out the flop. I needed an Ace on one of the remaining two cards or I was going to lose a $4k pot. The last two cards were inconsequential, Villain tabled his two 9’s, and took down the pot.

I immediately got up from the table to take a quick break. Intellectually, I knew I had made the right decision. Emotionally, I was pissed! That pot was nearly a month’s rent (San Francisco sucks sometimes). But the dissonance got me thinking: why the vastly different reactions from my thoughts vs my emotions? Why did I feel so shitty, even when I knew I had made the right decision? And what could that conflict teach me about the nature of mistakes and decision-making?

Why we screw things up

That we make mistakes isn’t in question. What’s amazing is that we don’t do it more often, given:

  1. We are imperfect beings, optimized through evolution for survival and reproduction

  2. We are living in a world with countless other imperfect beings

  3. The information coming to us from those other beings and the surrounding world is imperfect and incomplete

  4. We must make hundreds of decisions every day based on this imperfect information

Yet despite severe limitations that would cripple the world’s most powerful computers, most of us still manage to navigate the world successfully enough to survive, live on our own, hold down jobs, reproduce, and raise families that begin the cycle anew. How is that possible?

The answer is as incredible as it is simple: we guess.

Consider it from an evolutionary perspective. Take organisms with massive energy requirements (our bodies + brains) and put them in a world with finite resources (food). How does that organism survive? 

One option would be to create a perfect organism, one that takes in every sensory input and makes a perfectly logical decision from those inputs. The problem is that organism would require impractically large amounts of energy to survive and even more to reproduce. We can see to this a first approximation with supercomputers that use as much energy as a small town yet have less processing power than our brains which consume a million times less. 

The other option is to create an imperfect organism, one that relies on limited sensory inputs and infers successive approximations (guesses) based on those inputs. And what if instead of requiring conscious (and energy-expensive) thought to make those guesses, they could be done “in hardware” and carried out subconsciously, with additional interpretation and refinement as needed by the conscious engine. 

This process of refined, educated guesswork is how humans actually make decisions. Our brains “guess” at what they expect to find in the world based on current sensory inputs, and re-evaluate based on new data. Most of this happens subconsciously, with some additional processing at higher levels of thought. That this process is so primitive is partly why we’re prone to making so many mistakes. It also explains why we experience mismatches between the conclusions drawn by emotional and intellectual centers of our brains. It also gives us a blueprint for how to improve.

Categorizing our mistakes

Even though we are faced daily with an extremely large number of decisions that lead to an even greater number of potential outcomes, we can simplify our treatment of them by asking two simple questions: 1) did our decision make it more likely that we’d experience a bad outcome (a mistake of choice), and 2) was the actual outcome bad, relative to the outcome desired (a mistake of consequence).

Given that, we get the following combinations of decisions and outcomes:

I. Bad decision, bad outcome

II. Bad decision, good outcome 

III. Good decision, bad outcome

IV. Good decision, good outcome

For the purpose of this conversation, I’ll focus on those that could be potentially classified as mistakes (i.e. they have at least one “bad” element), which removes IV.

Most people focus exclusively on Type I mistakes, as they are comparatively easy to identify and learn from: “I drank expired milk, I got sick afterwards, ergo don’t drink expired milk.” While it’s important to evaluate and learn from Type I mistakes, their simplicity and ubiquity makes them ineffective as learning differentiators. That is, by focusing exclusively on Type I mistakes, you’re much less likely to identify opportunities to distinguish yourself from your competitors. This is true whether we’re talking about getting a job, securing a date with a crush, or winning a sporting competition. And yet, this is where the vast majority of people and companies focus their time and effort.

Type II mistakes are more difficult to identify than Type I, but as a result provide an opportunity for significant differentiation and learning. Since Type II mistakes result in a positive outcome (at least initially), people often fail to interpret them as mistakes. Humans, being guessing machines, tend to draw immediate connections between actions and subsequent events, no matter how tenuous the connection (e.g. I saw a black cat, I fell and sprained my ankle, ergo black cats are bad luck). 

Type II mistakes can also hijack our reward systems that drive habit-forming. Given the positive outcomes, Type II mistakes are more likely than Type I mistakes to lead to us forming bad habits that given enough time and sample size, exact long-term damage. Since breaking a habit is much harder than forming one, as anyone who’s tried to quit smoking can surely attest, fixing a Type II mistake requires significantly more vigilance to find, admit to, and address the initial error – something most people are unwilling or unable to do. 

Type III mistakes aren’t actually mistakes, but as a result of the bad outcome they’re often viewed as such. This misattribution has a tendency to result in errors of omission by causing people to neglect decisions that would be more likely to have a net positive effect. I can’t tell you how many times I’ve seen people at the blackjack table refuse to double on an 11 or hit on a 16 because it led to a bad outcome the time before, even though doing so has a positive return. 

Like Type II mistakes, Type III mistakes are more difficult to diagnose and resolve, and thus people and businesses can capitalize accordingly to gain an advantage. After all, cities like Las Vegas and Macau are built on avoiding Type III mistakes while exploiting people who make Type II ones, at margins of only a few percent.

Preventing future screw-ups

So, how is any of this useful?

First, our tendency towards mistakes illustrates why it’s critical to understand the “why” behind any decision. That is, to identify and understand the underlying factors and impact that our decisions have on particular outcomes. For example: in dating there are some universal truths that will benefit you at any stage in life, but just because it landed you a cute girl’s number in college doesn’t mean ripping a beer bong at age 40 is going to serve you well (far be it from me to discourage you though).

Second, you must continually evaluate and test the assumptions that drive your decision-making. This holds true even when – and especially when – things are going well. There’s a reason that there are endless mantras warning us against getting “too comfortable” with success, and that these mantras exist in folklore, modern texts, and business school case studies. No matter how accurate your initial assumption, at least some of the realities that underpin their success will change over time.

Lastly, you must get comfortable with embarrassment that comes from actively exploring your mistakes. No one likes being confronted with their mistakes; seeking them out to find hidden ones is doubly challenging. In a past leadership role I ran a “share your screw-up” session during our regular team meetings. During that session, each team member was encouraged to share a consequential mistake that they made during the week, after which the entire team would discuss the learnings from it. While initially uncomfortable, that these sessions were team-driven events, and allowed key learnings to be disseminating to everyone, made this a valuable exercise that people actively engaged with.

The result

After losing with the Aces, the break was exactly what I needed. It gave me time to let my emotions subside and re-evaluate my assumptions to consciously address any gaps in my reasoning. One thing I’ve learned from playing thousands of hours of poker that can be applied to anything in life: with any complex situation, you’ll always make a worse decision when you’re in a negative emotional state. This isn’t a personal shortcoming, it’s by (evolutionary) design.

Once my emotions calmed I bought back into the game and re-focused my energy on making the correct decisions, outcomes be damned. Several hands later I won a huge pot that brought me back to even. By the end of the session, I was solidly in the black and headed home – alas, not with a month’s rent, but with at least enough to keep the lights on until the next session.

Hope you enjoyed this week’s newsletter. As always, I welcome any feedback and ideas for future topics of conversation. 

Until next time,

Greg