Scientists tell each other a story that is half affectionate and half akin to a covert handshake. When someone brings up a difficult issue, someone else will shrug and remark, “We ran the Monte Carlos.” When you hear the phrase for the first time, it always sounds strange, as though a group of researchers spent a long weekend at a roulette wheel and returned with solutions. That is, in a sense, close to the reality.
In the most unlikely of circumstances, the Monte Carlo method was developed in 1946. Polish mathematician Stanislaw Ulam, who had fled Europe prior to the war, was recuperating from emergency brain surgery and was understandably concerned that his brain might never function normally again.
| Detail | Information |
|---|---|
| Method Name | Monte Carlo simulation |
| Coined By | Stanislaw Ulam (with John von Neumann, Nick Metropolis, Robert Richtmyer) |
| Year Conceived | 1946, Los Alamos National Laboratory |
| Original Purpose | Neutron diffusion calculations for the hydrogen bomb |
| Named After | The Monte Carlo casino in Monaco (Ulam’s gambling-loving uncle) |
| Type of Math | Stochastic / probabilistic simulation |
| First Major Use | Manhattan Project follow-up research |
| Modern Uses | Finance, AI, climate modeling, drug discovery, risk assessment |
| Key Requirement | Repeated random sampling within realistic input ranges |
| Cultural Footprint | Used by EPA, NASA, hedge funds, particle physics labs |
He sat playing solitaire, bored, exhausted, and a little scared. In between shuffles, he pondered what the chances were that a well-prepared hand would win. The combinatorics were enormous; there are more ways to sort 52 cards than there are atoms in the observable universe. So he tried a more straightforward approach. Simply count while dealing the hand repeatedly. It’s a tiny mental leap that only becomes apparent when someone makes it. However, it altered computing. John von Neumann was, as always, a few steps ahead when Ulam presented the concept to him. Around that time, Von Neumann once declared that he was “thinking about something much more important than bombs.” Computers were on his mind.
The team, along with Nick Metropolis and Robert Richtmyer at Los Alamos, modified Ulam’s solitaire trick to simulate the behavior of neutrons and, less romantically, the early hydrogen bomb. Metropolis, who was aware that Ulam had an uncle who was unable to avoid the Monte Carlo casino, came up with the name. The cultural imagination had not yet been engulfed by Las Vegas. The franchise was still at glamorous risk for Monaco.
It is almost embarrassingly easy to explain the technique itself. One simple equation can be used to solve deterministic systems, such as an Olympic shot-put leaving an athlete’s hand. Naturally, the world is rarely that submissive. In a city, smog spreads unevenly. The markets fluctuate. Different patients react to the same dosage in different ways. There isn’t a single answer for these systems, so you can’t compute one. Thus, you take a sample. You run the simulation thousands or millions of times, feed in random inputs within reasonable bounds, and let the overall shape of the results speak for itself.
Perhaps the most underappreciated feature of Monte Carlo is its patience. No sophisticated proof, no flash of theoretical genius. Simply keep repeating until the sound becomes a picture. Deterministic simulations always yield the same result, but stochastic ones don’t, and that’s the point, according to John Guttag of MIT.
These days, Monte Carlo simulations are quietly present in nearly everything that deals with uncertainty. In the 1990s, the EPA began to rely on them for risk assessments. At three in the morning, hedge funds operate them on delicate portfolios. They are used by particle physicists to pursue results that are elusive. Self-driving systems are operated by engineers on a billion-person scale. The same probabilistic playbook is used by some machine-learning models, which are the kind that power tools that people use without thinking.
Observing this gives me the impression that Ulam discovered something more significant than he first thought. In the end, a man dealing cards out of restlessness while recuperating from surgery gave the modern world a way to make sense of chaos. For a small collection of poems, it’s not bad.
