The rationality of irrationality in the Monty Hall problem

The rational solution of the Monty Hall problem unsettles many people. Most people, including the authors, think it feels wrong to switch the initial choice of one of the three doors, despite having fully accepted the mathematical proof for its superiority. Many people, if given the choice to switch, think the chances are fifty-fifty between their options, but still strongly prefer to stay with their initial choice. Is there some sense behind these irrational feelings? We entertain the possibility that intuition solves the problem of how to behave in a real game show, not in the abstract textbook version of the Monty Hall problem. A real showmaster sometimes plays evil, either to make the show more interesting, to save money, or because he is in a bad mood. A moody showmaster erases any information advantage the guest could extract by him opening other doors which drives the chance of the car being behind the chosen door towards fifty percent. Furthermore, the showmaster could try to read or manipulate the guest’s strategy to the guest’s disadvantage. Given this, the preference to stay with the initial choice turns out to be a very rational defense strategy of the show’s guest against the threat of being manipulated by its host. Thus, the intuitive feelings most people have about the Monty Hall problem coincide with what would be a rational strategy for a real-world game show. Although these investigations are mainly intended to be an entertaining mathematical commentary on an information-theoretic puzzle, they touch on interesting psychological questions.

💡 Research Summary

The paper revisits the classic Monty Hall puzzle, not to dispute the well‑known result that switching yields a 2/3 chance of winning versus a 1/3 chance for staying, but to ask why many people intuitively prefer to stay even after they understand the mathematics. The authors argue that the textbook formulation assumes a perfectly “fair” showmaster who always opens a goat‑containing door that the contestant did not pick. In real television game shows, however, the host may sometimes act “evil” – opening the contestant’s chosen door when it hides a goat, thereby forcing an immediate loss, or otherwise manipulating the information presented.

To capture this, they introduce a parameter p = P(evil), the prior probability that the host behaves maliciously. When the host is evil (probability p) the contestant’s chance of winning by staying is 1/3 (only when the initial pick was the car) and by switching is 0 (the host will always reveal the contestant’s door). When the host is fair (probability 1‑p) the usual 1/3 vs 2/3 split applies. Consequently the expected win probability for staying is always 1/3, while for switching it is (1‑p)·2/3. Thus switching is optimal only if p < ½; staying is optimal if p > ½; and the two strategies are equivalent at p = ½. The authors suggest that a rational host seeking to maximise audience tension while limiting costs would deliberately set p≈½, thereby erasing any informational advantage for the contestant.

The paper then applies Bayes’ theorem to the observation that the host opens “another” door (i.e., not the contestant’s). The posterior probability that the host is evil given this observation is P(evil|other) = p/(3‑2p). For p = ½ this posterior is ¼, meaning the host is still likely to be fair. More crucially, the posterior probability that the car lies behind the contestant’s original door becomes P(car|other) = (1‑2p)/3, which equals ½ when p = ½. This matches the common intuition that the remaining two doors each have a 50 % chance, even though the underlying model includes a potentially malicious host.

The authors further explore a “mind‑reading” host who can infer the contestant’s intended strategy (stay or switch) with probability q. If the host knows the contestant will stay, he can act fairly; if he knows the contestant will switch, he can act evil, driving the overall win rate down to ≤ 1/3. Conversely, a contestant can attempt to bluff, pretending to stay while actually planning to switch, but this carries the risk of being detected and punished.

In the final sections the paper connects these formal results to psychological literature. Human beings are evolutionarily tuned to detect deception; thus the intuition that “switching feels risky” can be seen as a protective bias against a potentially manipulative host. The authors cite the “insensitivity to prior probabilities” effect described by Tversky and Kahneman, noting that people often ignore the prior 1/3 chance of having initially chosen the car when presented with new evidence, leading them to assign a 50 % chance to each remaining option.

Overall, the paper concludes that the apparently irrational preference for staying is actually a rational defensive strategy when the contestant accounts for the possibility of a hostile or manipulative showmaster. The analysis blends information theory, game theory, and behavioral economics to explain why the public’s gut feeling aligns with a rational response to a more realistic, non‑idealized version of the Monty Hall game.