Monty Hall problem

Three doors. A prize behind one, nothing behind the other two. You pick a door, the host opens a different door to show it is empty, and then comes the offer: stick with your original pick, or switch to the other closed door? Play a good handful of rounds each way and let your running tally settle the argument. Before you start, it helps to know which side you are probably on: almost everyone's first instinct in this game is to stay with the door they originally picked. Across studies in several countries, roughly 80-90% of people stick on their first play — in the classic experiment only about 12% switched — even though switching is twice as likely to win.Convergent / replicated

The host's reveal looks like neutral information, as if the game had simply shrunk to a coin toss between two doors. It has not, because the host was never free to open just any door. Under the standard rules — the host knows where the prize is, always opens a losing door you did not pick, and always offers the swap — switching wins 2 out of 3 times. Your first pick keeps its original 1-in-3 chance, so the host's forced reveal funnels the remaining 2-in-3 onto the other closed door.Convergent / replicated Another way to feel it: switching loses only when your first pick was already right, and your first pick is right one time in three.

This puzzle earns its place as a demonstration that a reasoning error can be nearly universal and stubbornly resistant to expertise. This puzzle famously fooled experts, not just beginners. Statistician Steve Selvin posed and named it in two 1975 letters, and when Marilyn vos Savant published the switch answer in Parade in 1990 she received about 10,000 letters — many from PhD holders on mathematics and science department letterheads insisting she was wrong — and the mathematician Paul Erdos reportedly accepted the answer only after seeing a computer simulation.Historical / descriptive If the answer still feels wrong to you, you are in distinguished company — and the station above is your own small version of the simulation that convinced Erdos.

A tempting takeaway is that switching is simply the smart move whenever someone reveals an empty option. Careful. The 2-in-3 advantage is not a universal rule that switching is always better — it depends entirely on the host's behavior. If the host does not know where the prize is and merely happens to reveal a goat by luck, the two remaining doors really are 50/50, so the game demonstrates the value of switching only under the standard knowing-host rules.Convergent / replicated Our host is scripted to follow those standard rules exactly, and the game tells you so before you play. Nor do a few rounds of one famously counterintuitive puzzle say anything broader about your judgment; your session summary describes your moves here, nothing more.

Why does such a small puzzle defeat so many people? People go wrong here in predictable ways: they treat the two remaining doors as a 50/50 split and overlook that the host's move depends on their own choice. How the problem is presented matters a lot — when Krauss and Wang reframed it using counts of possible cases and the host's perspective, correct reasoning about the switch advantage rose from about 3% to nearly 40%, which still leaves most people fooled.Supported, with limits

One honest caveat about what practice can do. Playing the game repeatedly with feedback helps, but only partly. In Granberg and Brown's repeated rounds switching climbed from 12% to about 55%, and a computer simulation is what finally persuaded Erdos — yet reviews find people rarely reach the optimal always-switch strategy or a correct grasp of why it works, and in one head-to-head study humans failed to reach optimal play even with extensive training that pigeons mastered.Supported, with limits So treat a rising win rate here as progress, not as proof that the logic has clicked — the research suggests those two things can come apart.