Monty Hall Problem
mental-model
Source: Probability
Categories: mathematics-and-logicdecision-makingpsychology
Transfers
Three doors. One hides a car; two hide goats. You pick door 1. The host, who knows what is behind every door, opens door 3 to reveal a goat. Should you switch to door 2? Yes — switching wins two-thirds of the time. The problem is famous not because the math is hard (it is elementary) but because the correct answer violates strong intuition, and because even professional mathematicians initially got it wrong.
The Monty Hall problem is a mental model for recognizing when new information should change a decision — and for understanding why our brains resist the update.
Key structural parallels:
- Information asymmetry between revealer and chooser — the host knows what is behind every door; the contestant does not. When the host opens a door, this is not a random event — it is a constrained reveal. The host must show a goat, must avoid the contestant’s door, and must offer a switch. These constraints are what make the reveal informative. In decision-making, the model teaches that the value of new information depends on the informer’s knowledge and constraints, not just the information itself.
- Probability redistribution, not elimination — the naive intuition says: “Two doors left, 50/50 chance.” The correct reasoning says: your initial pick had a 1/3 chance; the other two doors collectively had 2/3. The host’s reveal eliminates one of those doors but concentrates its probability onto the remaining one. The 2/3 doesn’t disappear; it flows. This is the core transfer: when options are eliminated by a knowledgeable agent, the surviving options absorb the eliminated probability. In investing, this maps to updating valuations when a competitor exits a market.
- The stickiness of initial commitment — contestants who pick door 1 feel ownership of that choice. Switching feels like admitting a mistake, even though the initial choice was random and carried no information. The model exposes status quo bias: people anchor to their first decision and require overwhelming evidence to abandon it, even when the rational threshold for switching is easily met.
- The expert’s embarrassment — when Marilyn vos Savant published the correct solution in Parade magazine (1990), she received thousands of angry letters from PhD mathematicians insisting she was wrong. The model demonstrates that domain expertise does not protect against intuition failures, and that the strength of conviction is uncorrelated with correctness when the problem structure is counterintuitive.
Limits
- The host’s rules rarely apply — the problem’s elegant solution depends on the host’s constrained behavior: the host must open a door, must reveal a goat, and must offer a switch. In real decisions, the person revealing information may be acting randomly, strategically, or deceptively. If the host opens a door at random (and might reveal the car), the advantage of switching disappears. Applying the Monty Hall intuition to situations without a constrained, honest revealer is a misuse of the model.
- Single-play versus repeated-play — the 2/3 advantage is a frequency property: over many games, switching wins twice as often. In a single game, you either win or lose. For genuinely one-shot decisions (this acquisition, this hire, this surgery), the model’s prescriptive force is weaker because you cannot rely on frequency convergence.
- The model trains a heuristic that can misfire — “always switch when new information arrives” is not universally correct. The Monty Hall problem works because the new information is structured in a very specific way (constrained reveal by a knowledgeable agent). Generalizing to “always update, always switch” can produce fickleness rather than rationality.
- The problem’s fame overshadows its narrowness — the Monty Hall problem is the most widely known probability puzzle, which gives it outsized influence as a mental model. But its specific structure (three options, one knowledgeable revealer, one constrained reveal) is quite narrow. Most real-world updating problems have continuous evidence, multiple reveals, and uncertain informer reliability — situations where Bayes’ theorem is the right tool, not the Monty Hall shortcut.
Expressions
- “You should switch doors” — the canonical punchline, often delivered as a gotcha in probability discussions
- “The Monty Hall problem” — used as shorthand for any situation where intuition and probability diverge
- “But it’s fifty-fifty!” — the incorrect intuition, invoked as an example of base-rate failure
- “Let’s Make a Deal” — the game show that gave the problem its name, hosted by Monty Hall from 1963 to 1986
- “Ask Marilyn” — reference to the Parade magazine column where the controversy erupted in 1990
Origin Story
The problem is named after Monty Hall, host of the American game show Let’s Make a Deal (1963-1986), though Hall himself noted that the show’s actual format did not precisely match the problem’s stipulations. The mathematical puzzle was first posed by Steve Selvin in a 1975 letter to The American Statistician.
The problem became a cultural phenomenon in 1990 when Marilyn vos Savant answered it correctly in her Parade magazine column “Ask Marilyn.” Approximately 10,000 readers wrote in to object, including nearly 1,000 with PhDs. Paul Erdos, one of the most prolific mathematicians in history, reportedly refused to accept the solution until shown a computer simulation. The episode became a landmark case study in cognitive bias research and probability education, demonstrating that mathematical sophistication does not immunize against intuition failure.
References
- Selvin, Steve. “A Problem in Probability.” The American Statistician 29.1 (1975): 67
- vos Savant, Marilyn. “Ask Marilyn.” Parade, September 9, 1990
- Rosenhouse, Jason. The Monty Hall Problem: The Remarkable Story of Math’s Most Contentious Brain Teaser (2009)
- Granberg, Donald & Brown, Thad. “The Monty Hall Dilemma.” Personality and Social Psychology Bulletin 21.7 (1995): 711-723
Structural Neighbors
Entries from different domains that share structural shape. Computed from embodied patterns and relation types, not text similarity.
- Re-authoring (narrative/pattern)
- Creative Destruction (destruction/paradigm)
- Intoxication Is Getting Destroyed (destruction/metaphor)
- Inversion (geometry/mental-model)
- Character Is Conduct (theatrical-directing/mental-model)
- Problem Is A Target (target-practice/metaphor)
- Accidental Complexity (intellectual-inquiry/metaphor)
- The Shapeshifter (mythology/archetype)
Structural Tags
Patterns: splittingmatchingremoval
Relations: selectcause
Structure: transformation Level: generic
Contributors: agent:metaphorex-miner