The Monty Hall Problem: Why Switching Doors Is Mathematically Obvious (and Psychologically Impossible)

You're on a game show. Three doors. Behind one: a car. Behind the other two: goats. You pick door number one. The host, who knows what's behind every door, opens door number three to reveal a goat. He then asks: do you want to switch to door number two?

Photo by lil artsy on Pexels.

Most people say no. Most people are wrong.

This is the Monty Hall problem, named after the host of Let's Make a Deal. Marilyn vos Savant published the correct solution in her Parade magazine column in 1990 and received roughly 10,000 letters telling her she was an idiot. About a thousand of those came from people with PhDs. She was right. They were wrong. And the reason this problem breaks people's brains is worth understanding carefully.

What Your Gut Says

Your gut says: two doors remain, one has a car, so each door has a 50% chance. Switching gains you nothing. Stay or switch, the odds are identical.

This feels airtight. It's wrong.

When you picked door one at the start, you had a 1-in-3 chance of being correct. That means there was a 2-in-3 chance the car was behind one of the other two doors. Monty then does something crucial: he opens a door he knows has a goat. He never opens the winning door. He never opens your door.

His action carries information.

Why Switching Wins Two-Thirds of the Time

Walk through the three scenarios where the car sits behind each door:

Scenario A: Car is behind door 1 (your pick). Monty opens door 2 or 3. If you switch, you lose.

Scenario B: Car is behind door 2. You picked door 1. Monty must open door 3 (only goat option left). If you switch to door 2, you win.

Scenario C: Car is behind door 3. You picked door 1. Monty must open door 2. If you switch to door 3, you win.

Three equally likely scenarios. You win by switching in two of them. Stay and you win in one.

graph TD
    A[You pick Door 1] --> B{Car location?}
    B --> C[Door 1: Stay wins]
    B --> D[Door 2: Switch wins]
    B --> E[Door 3: Switch wins]
    C --> F[Switch loses 1/3]
    D --> G[Switch wins 1/3]
    E --> H[Switch wins 1/3]

Switching wins 2/3 of the time. Staying wins 1/3. The math is unambiguous.

The Real Problem: Monty Isn't Random

Here's where most explanations skip something important. The entire result depends on Monty's behavior being constrained. He always reveals a goat. He always reveals a door you didn't pick. He never reveals randomly.

If Monty opened doors randomly, sometimes accidentally revealing the car, the odds would shift to 50/50 after a goat reveal. The host's knowledge and intentionality are load-bearing parts of the probability calculation. Strip those out and you get a different problem entirely.

This is why the problem maps onto real analytical situations. Whenever a data source, an algorithm, or a human agent selects what information to show you based on prior knowledge, you're operating in Monty Hall territory. The selection process is encoding information. Ignoring that process means you're solving the wrong problem.

Why Smart People Still Get It Wrong

Two cognitive patterns conspire here. First: we treat probabilities as fixed properties of objects rather than as summaries of information states. Once Monty opens a door, we mentally reset the game to "two doors, fresh start" instead of carrying forward what we knew at the moment of our original pick.

Second: the uniform distribution feels like a default truth. Two options, no obvious reason to favor one, so 50/50. Psychologists call this proportionality bias. Our brains are bad at accepting that information structure, not just outcome count, shapes probability.

The mathematician Paul Erdős reportedly got this problem wrong on his first attempt. He only accepted the correct answer after seeing a computer simulation run thousands of trials. That's not embarrassing. That's evidence of how deeply this intuition is wired in.

What to Take Away

Run the simulation yourself. Write a quick script: stay 10,000 times, switch 10,000 times. Watch the win rates converge to 33% and 67%. The numbers are not subtle.

Then ask yourself how many decisions you make where a selecting agent with knowledge is curating your information, and you're treating the result as a neutral coin flip. Job offers. Market signals. Product recommendations. News feeds.

Monty Hall is everywhere. And almost nobody is switching doors.