Are You Smarter Than a Pigeon?

weirdly accurate.

    Are You Smarter Than a Pigeon?

    What is the Monty Hall problem, and can a pigeon beat you at it?

    This page is the Monty Hall problem played as a game: three doors, one pigeon behind them, and a host who opens an empty door before you decide whether to stay or switch. Play it and you find out what you do on instinct, whether that changes over ten rounds, and how you compare with the six pigeons and six students Herbranson and Schroeder tested in 2010. The right answer, and why it feels wrong, appear on your result.

    This test turns the puzzle into a learning curve, the way that experiment did. You play 10 quick rounds of the three-door game in two levels: a pigeon hides behind one door, you pick one, the host opens an empty door you didn’t pick, and you stay or switch before every door opens. A simulated rival pigeon starts from your first pick each round, so every point between you comes down to the choice you make after the host’s move. In level 1 it plays like the study’s pigeons on their first day; in level 2, like the same birds after a month of practice. A short stop between the levels gives the score, and one last round has 100 doors. Your result compares how often you switched in your first and last 5 rounds with those pigeons and with the six students in the same study, who played 200 trials each.

    It is not an intelligence test, and losing to a pigeon says nothing about your IQ. The pigeon and student figures are group averages from small lab samples — six birds and six students — who played far more rounds than you, and the students never saw a rival’s score, so the comparison is loose: your start and end are 5 rounds each, theirs a whole day or 50 trials. The closest study in length is Friedman’s 1998 experiment, in which 104 paid players played 10 rounds each; the result shows their switch rate beside your own. No study has published how individual people spread out on this repeated game, so there is no percentile. Five rounds is also a small sample: each round moves your rate by 20 points, so one choice either way can be luck.

    What the test measures

    One thing, round by round: whether you learn which choice pays, and how quickly.

    The task

    • The Monty Hall problem, the three-door choice named after the host of Let’s Make a Deal, played over and over as Herbranson and Schroeder (2010) gave it to pigeons and students
    • 10 rounds in two levels of 5: pick a door, the host opens an empty door you didn’t pick, then stay or switch; the pigeon is placed at random every round
    • A simulated rival pigeon that starts from your first pick: in level 1 it plays like the study’s pigeons on day 1, in level 2 like the same birds after a month of practice, with a live score and a stop between the levels
    • A final round with 100 doors, in which the host opens 98 empty ones

    Your result

    • The final score against the rival pigeon, and what it did differently
    • How often you switched in your first 5 and your last 5 rounds, beside the study’s pigeons on day 1 and day 30 and its students in their first and last 50 trials
    • How often you switched over your 10 rounds, beside Friedman’s 104 players, who also played 10 rounds each
    • From your own rounds: how many pigeons switching every time, or staying every time, would have found, and how often you won when you switched against when you stayed
    • The probability behind the puzzle, explained after you have played; one of four grades, set by how you played your last 5 rounds; no percentile, because no study has published one
    • A shareable card and a link that challenges a friend; the link carries only whether you switched in your last 10 rounds, in the part of the address after the #

    Before you start

    • About a minute; free, no sign-up
    • Works on a phone or a computer: tap the doors, or press 1, 2 and 3 on a keyboard
    • Sound is optional and can be switched off
    • The loft is drawn in 3D, so the browser needs WebGL switched on

    Monty Hall problem FAQ

    Is this test free? Do I need to sign up to see my result?

    Yes, it’s free, and there’s nothing to sign up for: no account, no email, no payment. Your result appears on this page the moment you finish, and you can share it or play again straight away.

    Should you switch doors in the Monty Hall problem?

    The pigeon test tells you on the result page, after you have played, because knowing first would change what you choose. Pick a door, watch the host open an empty one, then stay or switch, ten times over. Your result then shows how often each choice won in your own rounds, the exact odds of staying and of switching, and why two closed doors feel like a coin toss.

    Did pigeons really solve the Monty Hall problem?

    Pigeons did play the Monty Hall problem in a real study, and the pigeon test shows how they did on your result page, beside your own rounds. In Herbranson and Schroeder’s 2010 study in the Journal of Comparative Psychology, six pigeons played up to 100 trials a day for 30 days, with grain as the prize. Their choices on day 1 and day 30 set how the rival pigeon plays. Nobody claims the birds understood the probability; they learned from what happened.

    How often do people switch in the Monty Hall problem?

    The pigeon test shows that on your result page, after your own rounds, because it is the number your choices are measured against. The figures come from three sets of studies: Granberg and Brown’s 1995 experiments, Friedman’s 1998 paid experiment with 104 players over 10 rounds each, and the six students in Herbranson and Schroeder’s 2010 study, who played 200 trials. The result also names the habit Herbranson (2012) saw in those students.

    How does the pigeon test score you?

    The pigeon test grades you on how you chose in your last 5 rounds, not on how many pigeons you found, because finding the pigeon is partly luck. The rival pigeon starts from your first pick every round, so the score between you depends only on staying or switching. Five rounds is a small sample, so the result also shows how many pigeons switching every time, or staying every time, would have found in the exact rounds you played.

    Is “Are you smarter than a pigeon?” an IQ test?

    No. The pigeon test is a game, not an intelligence test, and it measures one thing: how your choices change in a repeated three-door game. Losing to a pigeon says nothing about your IQ, and the pigeons in the 2010 study weren’t smarter than people; the result explains how the birds and the students differed. The comparison figures come from six birds and six students, so read them loosely.

    Who came up with the Monty Hall problem?

    The Monty Hall problem is named after Monty Hall, host of the American game show Let’s Make a Deal, and the statistician Steve Selvin is usually credited with posing it in The American Statistician in 1975. It became famous in 1990, when Marilyn vos Savant answered it in her “Ask Marilyn” column in Parade. By her own estimate she received about 10,000 letters, most saying she was wrong, close to 1,000 of them signed by people with PhDs. Her answer, and who turned out to be right, are on the result sheet.

    How long does the pigeon test take, and does it work on a phone?

    The pigeon test takes about a minute: two levels of 5 rounds with a short stop between them, then one 100-door round. It runs in the browser on a phone or a computer with 3D graphics (WebGL) switched on; tap the doors or press 1–3. It is free and needs no sign-up.

    Where the numbers come from

    The pigeon and student figures are group averages from Herbranson & Schroeder (2010): six pigeons over 30 days of up to 100 trials, and six students over 200 trials; the paper reports only their first and last points. Your rival pigeon is simulated at the pigeons’ day-1 switch rate for the first 5 rounds and their day-30 rate after that. Friedman (1998) supplies the same-length comparison: 104 people, 10 rounds each. There is no percentile, because no study has published individual results for this repeated game, and the site does not publish its own players’ data. The letter counts are vos Savant’s own estimates.

    How the score works

    Switch rate

    The share of rounds in which you switched, counted in blocks of 5, one per level. Your grade uses the last block, because that shows what you learned; the 100-door round is reported on its own and not counted.

    The rival pigeon

    Each round the rival starts from your first pick. In the first 5 rounds it switches with the day-1 probability of the pigeons in Herbranson and Schroeder (2010); after that with their day-30 probability. The result shows both rates. It is simulated, not a recording of a real bird.

    What switching would have found

    From the doors you were actually dealt, the result counts how many pigeons switching every time, or staying every time, would have found, so you can compare your choices with the same luck.