secretary problem

when should you stop looking?

You're hiring. N candidates arrive in random order. After each interview, you must hire or pass — no callbacks. The optimal strategy: reject the first 37%, then hire the next candidate who's better than everyone you've seen.

optimal threshold = N/e ≈ 37%   P(best) ≈ 1/e ≈ 36.8%

This is the optimal stopping problem. The 1/e rule maximizes your chance of picking the single best candidate. Play the game below, then run the simulation to see the peak at 37%.

references

Ferguson. "Who solved the secretary problem?" Statistical Science, 1989.

Bearden. "A new secretary problem with rank-based selection." Journal of Mathematical Psychology, 2006.

interactive
press "new game" to start
games 0
picked best 0
success rate

monte carlo simulation

For each rejection threshold (0%–100%), simulate thousands of hiring rounds. The peak reveals the optimal stopping point.

trials per point 0
peak threshold
peak success