How to better define your design problems — or why Thanos didn’t double resources
Spoilers for Infinity War, but not Endgame
Avengers: Infinity War features a powerful super villain: Thanos. The central plot follows him on his quest to kill half the population of the universe. He isn’t doing this for evil reasons though, his stated goal is to “restore balance” to the universe. Thanos sees overpopulation as a threat and believes the solution is to kill half of all creatures so the other half can live better on the available resources.
This brings up an important question: if he’s all powerful, why not grow more food and double the resources instead?
When we face a design problem, we often spend most of our time going from problem to solution. But could we be ignoring better options because we haven’t defined our problem well in the first place?
A solution too early could interfere with how you think about your problem
Thanos first considered his own home: Titan. He saw Titan struggle to provide for its populace and blamed overpopulation. He proposed to his government that the solution was in killing half of their people. This was rejected as genocide and not acted upon. As he predicted though, the planet did suffer and his civilisation collapsed. That led him to conclude that his solution was the only way forward.
Since Thanos was already convinced that he had a great solution, he interpreted the results only from his perspective. Although there is no evidence that a smaller population would fare better, he is now certain his solution is perfect and applicable to the entire universe.
This is often referred to as a confirmation bias. The reasons for the collapse may have been more complex than “overpopulation”. His solution may not have worked as well as he’d hoped. But since the end result was just as he expected, it confirmed his world view. He was now even more convinced that killing half of everyone was the right thing to do.
Thanos has a solution looking for a problem. His stated goal was to “restore balance”, but because his solution only involves killing, he starts to define his problem as “too many people”. From here, his attention is on how to kill people efficiently.
Limitations of building a partial solution
When a problem has been defined by a solution, rather than the other way around, it can limit its effectiveness. The solution may not be holistic, but since the problem was rewritten to fit, it seems perfect. We’re moving to production half-baked leading to a limited scope of work. The team we build and the processes we develop, will now only answer the poorly defined problem.
In Thanos’ case, he has moved his focus to destroying entire populations. He has hired staff based on ability to kill. He has invested in a warship. The only task he is prepped to handle, is killing populations. Each new planet he visits could have individual concerns, implementation could be localised, but all he has is a killing machine.
“To the man who only has a hammer, everything he encounters begins to look like a nail.” —Abraham Maslow
As an example, let’s take Earth: killing 3.6 billion people would take us back to our population in 1972. That gives us only 50 years before we’re right back to where we are now. Without additional support systems and planning, nothing has truly been improved. Thanos built none of this infrastructure to aid the remaining populace. Because his process only involved reducing the population, sustaining the rest is outside his scope of work and he built nothing toward that end.
Poorly defined KPIs can kill your strategy
An important thing to note here is his metric for success—his Key Performance Indicators. The only one he has set for himself is to kill half the creatures—half of all the creatures. So animals as a food source, those that help pollination, or assist in cultivation, are killed as well. Both creatures populations generating resources and those consuming them are cut in half. He is reducing the very resources people are meant to survive on.
This seems counter to his stated goal of “restoring balance”. If you have an unbalanced scale and halve both sides, the scale remains unbalanced.
A better metric might be the ratio between available resources and mouths to feed. It might be a planet’s more efficient use of energy. Or reduced food wastage. He is disregarding many other solutions that could result in a better final outcome. A narrow problem definition produced a poorly defined KPI incentivising killing over balance.
We’ve seen this happen with journalism and click-bait. When we measure an article’s success based on number of clicks rather than knowledge disbursed, it becomes more important that someone click a link than read it. The value of the headline far outweighs the substance of the piece.
Scaling solutions amplifies imperfections
Thanos realised killing creatures one planet at a time was tedious and unnecessarily violent. He had no personal vendetta, no need for bloodshed. The cleaner implementation to him, was to use the power of the Infinity Stones. These stones would allow him to scale his solution across the entire universe with one snap of his fingers. He could turn half of all creatures to dust in an instant.
The stones are not a one-trick pony however. This is a new tool to Thanos. A tool that can do vastly more than his army. They grant him the power to do practically anything he wants — control living and dead souls, alter time, omnipresence across the universe, influence people’s thoughts and dreams, manipulate energy. The Reality stone alone can grant any wish even in direct contradiction with scientific law.
Infinity Gems on Wikipedia
https://en.wikipedia.org/wiki/Infinity_Gems
That’s a lot of power to direct towards a limited, untested solution. If Thanos had focussed on his initial problem, he could easily attempt many alternative solutions and even go back and forth in time to test them.
When all you want to do is hit nails, everything you have can be a hammer. —Me
When we scale, we take our limitations with us. New software, new hires, new processes, all inherit the same focus areas we had already defined for ourselves. If our solutions are short-sighted, better systems will only hit those limits faster.
Iterative design and marginal gains
Thanos put everything he had into one magic-bullet solution. Instead of that singular focus, what if he spread his gains across many solutions?
In 2003, British Cycling hired Dave Brailsford as performance director. Toward the big goal of making athletes faster, he made tiny improvements everywhere. From helmets, to seat shape, fabrics, body temperature, tires, sleeping arrangements… he focused on improving little things by 1%. All that then added up in a strategy he referred to as “the aggregation of marginal gains”.
Within a decade, Britain went from winning one olympic medal in a hundred years, to taking 60% of gold medals and setting seven world records.
Olympics cycling: Marginal gains underpin Team GB dominance
https://www.bbc.com/sport/olympics/19174302
Big redesigns are tempting. Dribbble is full of visual updates to popular sites. But a functioning product often needs iterative improvement—small changes that add up over time. With a clear problem definition and well structured KPIs, each improvement is a step toward a common goal.
Conclusion: Designing good problems
Simply by framing his problem too narrowly, Thanos limited his solutions and destroyed lives instead of improving them. The way you ask your question has a large impact on how you solve it.
- Define your original problem broadly
Give yourself space to explore solutions tangential to your primary goal
Don’t let your solution define your problem - Break up your KPIs
Incentivise a variety of positive outcomes
Everything is focussed on improving metrics, make sure they’re right - Reconsider your problems when you scale
A new toolset could mean a larger pool of solutions
Not reviewing goals could mean wasting the potential of the new - Iterate for marginal gains
Small improvements add up quickly
Big redesigns aren’t the only path to better
