One Empty Locker, Two Inspections

A team of N−1 players shares N lockers: N−1 hold labelled objects (one each, matching a player) and exactly one locker is empty. Each player may open up to two lockers, choosing the second based on what the first revealed. The team wins only if every player finds their own object. This toy lets you simulate three strategies and watch the empirical win-rate converge on the paper's closed form, IN / N! — the count of involutions of N elements, over N factorial.

arXiv:2608.13074 · "One Empty Locker and Two Inspections: An Exact Optimal Team Strategy" β†—

Setup

Choose the team size. Lockers = N, players = N−1, exactly one locker is empty.
10
Players
9
I(N) involutions
9,496
N!
3,628,800
Theory I(N)/N!
0.26%

Watch One Trial

Pick a strategy and play a single randomized trial, player by player, to see exactly how the two-inspection rule works.
Press "Play animated trial" to begin.

Live Convergence β€” Monte Carlo vs Theory

Run trials in bulk. Each strategy's empirical win rate is tracked live against the closed-form theoretical rate (dashed).
Random guess
0.0%
0 trials
Naive pointer
0.0%
0 trials
Optimal pointer
0.0%
theory 0.0%

The Big Reveal

Three exact, computable closed forms at the current N β€” no simulation needed to know these numbers, only to watch them converge.
1 inspection each (no coordination trick possible)
P = 1 / N!
4.2%
2 random guesses each
P = (2/N)^(N−1)
0.0%
2 inspections, optimal pointer strategy
P = I(N) / N!
41.7%