Proby, the Probastack mascotprobastack

The Monty Hall Trap

Play the game, trust your gut, then run a thousand trials and watch your intuition fall apart. A hands-on tour of why switching wins two-thirds of the time.

Beginner11 min read·Conditional Probability · Bayesian Thinking · Sample Space

The hook

You’re on a game show. Three doors. Behind one is a car; behind the other two, goats. You point at a door — and then the host, who knows where the car is, swings open a different door to reveal a goat.

Now the offer: stick with your door, or switch to the last unopened one. Most people feel it’s a coin flip — two doors left, 50/50, why bother moving? Hold onto that feeling. We’re about to test it.

On the big screen

You may have seen this exact puzzle before. In the movie 21, an MIT professor springs it on a student to test whether he’ll think or just guess. Watch how fast intuition and probability part ways:

From the film 21 (2008) — the Monty Hall problem, dramatized.

Play it yourself

Don’t take anyone’s word for it — play a few rounds. Pick a door, see the host reveal a goat, then choose. Keep an eye on the two tallies at the bottom.

Pick a door.

When you stayed
0/0 won
When you switched
0/0 won
Play 10–15 rounds. Try staying for a streak, then switching for a streak.

Make a prediction

Before the math: if you played a thousand games always switching, what win rate would you expect? Commit to a number in your head.

Run a thousand

0 trials
Always stay
0 / 0
0%
Always switch
0 / 0
0%

Each trial draws a random car and a random first pick, then scores both strategies. Run a thousand and watch the rates settle near 1⁄3 and 2⁄3.

The law of large numbers doing its job, live.

The more trials you run, the tighter the rates lock onto 13\tfrac{1}{3} and 23\tfrac{2}{3}. So where does the missing intuition go? Why isn’t it a coin flip?

The reveal

The trick is that the host isn’t opening a door at random — he knows where the car is, and he’ll only ever reveal a goat.

Your first pick was right 13\tfrac{1}{3} of the time and wrong 23\tfrac{2}{3} of the time. The host’s reveal doesn’t change that — but it quietly sweeps the entire 23\tfrac{2}{3} of “you were wrong” onto the single remaining door. Look at all three equally likely cases:

Case 1 · 1⁄3
🚗

You first picked the car

Switching loses
Case 2 · 1⁄3
🐐

You first picked a goat

Switching WINS
Case 3 · 1⁄3
🐐

You first picked the other goat

Switching WINS

Switching turns every “you first grabbed a goat” into a win — and that happens twice as often as grabbing the car. The host hands you the 23\tfrac{2}{3} he was holding.

The math, gently

Let’s put numbers on the gut feeling. When you first point at a door, you’re right with probability 13\tfrac{1}{3} and wrong with probability 23\tfrac{2}{3}. That split is fixed the instant you choose — before the host touches anything.

Now the “always switch” rule: switching wins exactly when your first pick was a goat. That happens 23\tfrac{2}{3} of the time. So switching wins 23\tfrac{2}{3} and staying wins 13\tfrac{1}{3}— they’re just the two outcomes of that very first guess, seen from opposite sides.

Check yourself

Suppose there are 100 doors. You pick one, and the host opens 98 goats, leaving yours and one other. Switch or stay?

What is it about the host’s action that makes the reveal informative?

Now suppose the host opened a door at RANDOM and just happened to reveal a goat. Given that, what are your odds if you switch?