The 100 Prisoners Box Bet Explained: A Classic Probability Puzzle and Optimal Strategy

Quick take

This video presents a high-stakes probability puzzle: 100 players, 100 boxes, one bill per person, and each player may peek into only 50 boxes. To win, every player must find their own bill. A pre-game strategy is allowed, but no communication afterward.

The core idea is the cycle-following strategy, where players start with the box matching their own number and follow the numbers they find to successive boxes. If any cycle in the permutation exceeds 50, the group loses. The remarkable result is that the overall probability of all players succeeding is about 31% for 100 players. The video walks through the reasoning and then invites you to explore whether you should take the bet.

• Best strategy relies on permutation cycles
• All players must succeed for a payout
• Probability of success for 100 players is around 31%
• Understanding the implications of cycle lengths clarifies the puzzle

Overview

The video adapts a well known probability puzzle into a betting scenario. A group of 100 participants lines up, each takes a dollar bill labeled with a unique position, and places the bill in a box among 100 identical boxes. In a separate room, the boxes are shuffled, and each player, one at a time, may look into up to half the boxes (50 boxes) to locate their own bill. If every participant finds their own bill, the group wins a payout; if any one person fails, the entire stake is kept. Before entering the room, players may plan a strategy together, but there is no communication after anyone begins the search.

The Challenge and Rules

The task is essentially a coordination problem under strict local constraints. Each person must locate their own bill using at most 50 box inspections, with the added twist that the room is reset after each turn and no information can be shared between players after they leave. The payoff is high but the event of success depends on the collective performance, not any single individual outcome. The question posed is whether to accept the bet given the odds and the required universal success.

The Cycle-Following Strategy

The canonical solution uses a permutation view of box contents. Each box contains a number that corresponds to a person. If a player starts with the box whose number equals their own, they read the number inside and then jump to the box with that number, continuing this cycle until either they locate their own bill or exhaust 50 boxes. The strategy hinges on the structure of the permutation: if all cycles in the random permutation have length at most 50, then every player will eventually find their own bill within 50 steps. If any cycle is longer than 50, that cycle’s members will fail to find their bills within the limit, causing a loss for the entire group.

Why the 31% Figure Emerges

In the standard formulation with N players and N/2 allowed openings, the probability that all players succeed equals the probability that the random permutation on N elements has no cycle longer than N/2. For N = 100, this probability is surprisingly robust at about 31%. The explanation rests on the distribution of cycle lengths in random permutations: roughly half of all permutations have a longest cycle length less than or equal to N/2, and the cycle structure in large permutations yields the 31% figure as a natural threshold for universal success under the half-box constraint. Thus, despite the high stakes, there is a nontrivial chance of winning when everyone adheres to the cycle-following method.

Extensions and Takeaways

Beyond the specific numbers, the problem showcases a broader principle: local optimization under global success criteria can lead to counterintuitive results. The cycle-following strategy is elegant because it requires no inter-player communication during the challenge and leverages the inherent properties of permutations. For learners, the puzzle highlights how probability theory, combinatorics, and graph-like representations of permutations intersect to yield nonobvious insights. The video invites viewers to reflect on whether to accept the bet given the calculated odds and to consider how similar cycle-based strategies might apply to real-world coordination problems with limited information sharing.