Skip to main content
Category: Paradoxes
Type: Logic and probability paradox
Origin: Based on the TV game show “Let’s Make a Deal”; formal analysis discussed by Steve Selvin (1975), popularized by Marilyn vos Savant (1990)
Also known as: Three-door problem
Quick Answer — In the Monty Hall Problem, you should switch doors after the host reveals a goat. Switching wins with probability 2/3, while staying wins with 1/3, because the host’s action is informative rather than random.

​
What is Monty Hall Problem?

The Monty Hall Problem asks whether you should switch your initial choice when one losing option is revealed by a host who knows where the prize is.
The host’s constrained reveal transfers probability mass to the remaining unopened door.
Although many people feel “two doors means 50-50,” that intuition ignores information structure. The puzzle is a practical gateway to Bayesian Thinking, Expected Value, and Simpson’s Paradox where conditional information changes conclusions.

​
Monty Hall Problem in 3 Depths

  • Beginner: First choice is 1/3 likely correct; switching captures the other 2/3.
  • Practitioner: Always model what information the process reveals and what it hides.
  • Advanced: Equivalent formulations (many doors, simulation, Bayesian updates) all converge to switching advantage.

​
Origin

The puzzle derives from the U.S. game show “Let’s Make a Deal,” hosted by Monty Hall. Mathematical discussion appeared in statistical correspondence, including Steve Selvin’s 1975 treatment. It became globally famous after Marilyn vos Savant’s 1990 column advised switching and triggered large public debate, including criticism from many technically trained readers. Subsequent formal proofs and simulations repeatedly confirmed the same result: switching yields a 2/3 win rate under the standard rules.

​
Key Points

Monty Hall is not a trick question; it is a conditional probability lesson.
1

Initial probabilities do not reset

Your first pick has 1/3 chance of being correct and remains 1/3 unless new randomization changes that fact.
2

Host behavior is constrained

The host always opens a goat door and never opens your chosen door; this is informative.
3

Switching captures the complement event

Switching effectively bets that your first choice was wrong, which occurs with probability 2/3.
4

Simulations reinforce understanding

Running thousands of trials quickly converges near 66.7% win for switching and 33.3% for staying.

​
Applications

The same logic appears in real decisions where reveal rules are non-random.

Product Experiments

Interpret metric shifts based on how data was filtered or revealed, not just final counts.

Fraud and Security Investigation

Update suspicion based on what an attacker can and cannot choose to expose.

Negotiation and Bidding

Opponent disclosures are often strategic constraints, not neutral signals.

Decision Education

The puzzle is a compact way to train teams on conditional probability and process-aware reasoning.

​
Case Study

After Marilyn vos Savant’s 1990 column in Parade magazine, tens of thousands of letters challenged the switching answer. This public controversy became a real-world experiment in statistical intuition: many readers, including professionals, insisted the probability was 1/2 after one door opened. A measurable indicator came from repeated simulations and classroom trials. Across large trial counts, switching consistently produced win rates near 66.7%, while staying stayed near 33.3%. The debate showed that probabilistic errors can persist even among educated audiences when process constraints are overlooked.

​
Boundaries and Failure Modes

The 2/3 result depends on specific assumptions.
  • Host-rule mismatch: If host behavior differs (e.g., sometimes opens prize), probabilities change.
  • Ambiguous reveal process: If you do not know reveal rules, you cannot apply the standard result directly.
  • Over-transfer: Not every “new information” scenario behaves like Monty Hall; structure matters.

​
Common Misconceptions

The puzzle is often misunderstood because people compress it into a visual symmetry argument.
Correction: Equal count does not imply equal probability when one door is opened under constraints.
Correction: Switching changes your strategy to exploit the original 2/3 probability that the first choice was wrong.
Correction: The host’s rule-based action is precisely the source of additional information.
Monty Hall belongs to a family of conditional-information problems.

Bayesian Thinking

Updates beliefs when new evidence arrives under known process rules.

Expected Value

Converts strategy differences into long-run payoff comparisons.

Simpson's Paradox

Shows how aggregation without context can reverse conclusions.

​
One-Line Takeaway

Monty Hall teaches that good decisions depend on process-aware probabilities, not just how options look after information is revealed.