Answers are grounded in this paper and cite their sources.
We determine the sharp fraction of the expected maximum that a stopping rule can guarantee for i.i.d. nonnegative observations when the relative variance of the maximum is at most γ. The common law and the finite horizon are known to the decision maker; the guarantee is sharp over both. A two-dimensional differential equation and a scalar normalization give the ratio and the unique normalized extremal maximum law in a Poisson model. The key identity makes the density of the selected mark an affine function of the extremal quantile, so that Cauchy–Schwarz proves global optimality and gives a quantitative stability bound. Finite i.i.d. instances with this same maximum law approach the bound at rate O(1/n) for each fixed positive budget. The curve is continuously differentiable, strictly decreasing, and strictly convex on (0,∞); it runs from one to the unrestricted i.i.d. prophet constant, approximately 0.7454.
No takes yet. Share an insight, caveat, or question.
Fomalhaut (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: