Invert, always invert
Let’s say you need to solve a problem. Or make a decision. Maybe a project at work, advancing your career or a personal problem. How do you approach it? Follow your first idea? Do research? Ask someone for advice? Brainstorm? I could extend this list forever, and it would always remain the same: You think about the ideal outcome with an infinite number of options at hand.
Learning from Charlie Munger
But how do you know your decision is the right one? What is even the ideal outcome? What if you missed something and everything goes to hell? One of my all time heros, Charlie Munger — vice president of Berkshire Hathaway and longtime partner of Warren Buffet — proposes a different way to make decisions: Invert the problem. One day, while binge-watching recordings of old Berkshire Hathaway shareholder meetings (yeah, that’s how I spent my free time), I saw him talking about it and it immediately clicked for me:
As a young man, Charlie was a weather forecaster for the air corps in WW2. When spending all night in hangars drawing colors on weather maps, he always asked himself one question: “How could I most easily kill our pilots?”. That’s not a question most people would ask themselves, but it was easy to answer. One could either route the pilots into bad weather so that the planes would freeze or one could route them somewhere they would run out of gas and had to land in enemy’s land. For every mission, he fanatically tried to avoid those two basic hazards and as a consequence achieved superior results.
Charlie is also known for a more comedic version of this which is often quoted by Warren Buffet: “All I want to know in life is were I’ll die and then never go there.”
The concept of inverting the problem is counter intuitive to most people. We are usually told to be active and “strive for the best”. Inverting the problem on the other hand requires thinking about strategies for worst case outcomes and then simply preventing those. This seems passive, which generally has a negative connotation.
I believe there are two reasons why problem inversion is a useful trick. First, people consistently underestimate worst case outcomes and their probabilities. This may be due to ignorance or overconfidence bias.
Second, what makes inverting problems so powerful is that it decreases the problem space. By outlining what not to do (or actions to avoid certain outcomes), the set of strategies with a high probability for success to choose from gets limited.
Ever made a mistake?
Have you ever done something stupid? Something that you would clearly refer to as a mistake? Of course you have. And most likely you have come into similar situations again and immediately thought about this mistake you made last time. If it’s been a significant error, you surely did not fuck up the same way again and are happier with the outcome this time. See what I’m saying? You subconsciously inverted the problem. Learning from our mistakes and not making the same ones twice is not just common advice. It is problem inversion through experience. But you don’t have to wait for a mistake. Invert the problem in your head before, and avoid it altogether.
The prime example of problem inversion: Behavioral finance
The perfect example for the effectiveness of problem inversion is retail investing. People who set up an automated savings plan for an index fund outperform almost anyone else. How can such a simple strategy work so well? Invert the problem: How could you best avoid creating personal wealth?
You could spend all your income and not save a penny. Or, when investing, you could put all the eggs in one basket that eventually turns out to lose. If you avoid those two basic, but very common, mistakes by (1) forcing yourself to save money by setting up an automated savings plan and (2) by diversifying through a global index fund, you will be very well off.
When I read The Intelligent Investor by Ben Graham (“The best book about investing ever written” — Warren Buffet) a few years ago, it struck me that the book only had very few chapters with concrete advice for action. Most of its chapters described which mistakes to avoid. I immediately had to think about this when I listened to Charlie’s story above.
Problem inversion applied at work
Around nine months ago, we worked on a relaunch of some of our core insurance products at Getsafe. It took us forever and caused a fair bit of internal debate and frustration. How could something this simple take so long? Once the project was done and we prepared for the next planning’s cycle, we made two important decisions: (1) We formed a dedicated engineering squad (with me in the product manager role) whose sole purpose is to “make launching insurance products faster” and (2) dedicate a good part of my personal time to improve the processes around this.
In the early weeks of this phase when I needed to steer the direction of this new team’s focus and kick off the development of another new insurance product in parallel, I tried to actively invert the problem(s) all the time. Instead of going from detail to detail, I asked myself: “How could I most effectively delay the development of these new products so that nothing will ever get done?”.
One answer I came up with surprised my engineering-biased mind: The most efficient way to not get anything done would be to ensure that no one is aligned on the content, meaning and version of what actually needs to be done. Turns out we were very good at that! Changing decisions a lot, communicating those only through Slack DMs, having multiple versions of the same documents floating around, archived in different locations, using different terms for the same thing and vice versa across contributors and so on …
As a consequence, one of the first things I did was to establish a standardized wiki structure with predefined sections and formatting to document all major aspects of building a new insurance product. At the same time, we aligned on a standardized set of patterns for price models — one of the major components of every new insurance product — and derived a documentation template and later on a DSL for writing the code implementing it. This ensured that from actuary to engineer, everyone used the same structure and also the same ubiquitous language for its elements, all documented in a single place that’s accessible to everyone.
A whole set of problems we used to have in previous projects almost disappeared. We got very positive feedback from across contributors and project execution was measurably faster.
Stoicism and inversion of personal happiness
I’ve enjoyed reading a fair share of the stoic classics over the years and I tend to think about those every now and then when having a hard time. One day, my brain did its magic and I wondered: Is Charlie Munger a stoic? Did the ancient stoics consciously apply problem inversion?
So I asked myself: How could I most easily make myself miserable?
- Spending most of my day time on something (= a job) I hate.
- Be around people that pull me down and make me miserable.
- Get sick.
Consequently, life can’t be that bad when:
- Choosing an occupation and a team to work with that I’m passionate about.
- Surrounding myself with positively minded people of high integrity, ideally people better than me.
- Work out and eat healthy.
That’s quite a contrast to most people’s very concrete life plans of a specific job, a certain salary or idealized expectations of their family, a relationship and their friends. Do those expectations and plans lead to a better life though? Probably not. I know plenty of people (including myself sometimes) who don’t follow those three points, even though they are so simple. It is a challenge, but also not that hard.
So the next time when don’t know what to do: Invert, always invert.
