4.4/5 - (58 votes)

In 1921, a mathematician proved that someone wandering completely at random through a city will almost surely end up back where they started, but a bird doing the same thing in open air may never make it home.

A walk that starts with a coin toss

The idea behind it is called a random walk, and it is one of the simplest objects in all of probability. Put a marker at zero on a number line. Flip a coin. Heads, it moves one step right. Tails, one step left.

That is it. No strategy, no memory, no destination. Each step is a fresh 50/50 gamble.

After five flips giving three heads and two tails, the marker lands exactly on 1. And here is the surprising part: ten completely different sequences of flips lead to that very same spot.

The name itself is younger than you might think. British statistician Karl Pearson, the founder of modern biometrics, introduced the term “random walk” in the journal Nature in 1905.

Why you never drift as far as you think

Intuition says a random wanderer should either stay close to home or drift off far away. Neither is quite right.

The typical distance from the starting point grows like the square root of the number of steps. In plain terms, after 100 random steps you are usually only around 10 steps from where you began. After 10,000 steps, roughly 100.

On a simple line, something even stranger happens. The walk will eventually cross every single point an infinite number of times. Mathematicians call this recurrence, and it means nothing on the line is ever truly left behind.

See also  The BMI was invented by an astronomer, and he never meant it for patients

The drunk man and the drunk bird

Now picture a person lost in a city laid out as an endless square grid. At every intersection, they pick one of four directions (north, south, east or west) with exactly equal odds.

In 1921, Hungarian-American mathematician George Pólya proved that this wanderer will almost surely return to their starting corner, sooner or later. In probability, “almost surely” is as close to a guarantee as math ever gets.

Then add a third dimension, and the guarantee collapses. The chance of ever getting back to the origin drops to about 34%, and it keeps shrinking as more dimensions are added.

Japanese-American mathematician Shizuo Kakutani summed it up in one of the most quoted lines in probability: “A drunk man will find his way home, but a drunk bird may get lost forever.”

The myth that every random path comes back

It is tempting to believe that randomness always evens out, and that any wanderer given enough time must stumble back to the start. That is true in one and two dimensions. In three, it is simply false.

The same dimension trick shows up elsewhere. In 1960, Paul Erdős and Samuel James Taylor proved that two completely independent random walks will cross each other infinitely many times, but only in four dimensions or fewer.

Another common mix-up is treating the random walk and the Wiener process (the continuous model behind Brownian motion) as the same thing. They are not. A random walk moves in separate, distinct steps, while a Wiener process path is a genuine fractal. The walk only turns into the Wiener process when its steps are shrunk down in a very specific mathematical way.

See also  A 99 percent accurate medical test can still be wrong most of the time

That fractal is remarkably precise. In 2000, Lawler, Schramm and Werner proved that the outer boundary of such a path has a fractal dimension of exactly 4/3.

Why a fair game still ruins the gambler

Random walks are not just a curiosity about lost birds. They also explain one of the harshest truths about gambling.

Imagine a player with a limited bankroll betting against a bank with unlimited money, in a perfectly fair game. Every bet is a coin flip, so the player’s fortune follows a random walk. And since a walk on a line eventually visits every point, it will sooner or later hit zero.

No rigged odds are needed. A fair game and a finite wallet are enough to guarantee ruin. Mathematicians call it the gambler’s ruin problem.

A question of how much room you have

From a coin flip on a number line to a bird lost in the sky, the same simple rule produces wildly different outcomes depending on how much space it has to roam. Sometimes the most important number in a problem is not the odds at all, but the number of dimensions you happen to live in.