A truncated play-the-winner procedure for selecting the best of $ k geqq 3 $ binomial populations is proposed. Sampling always terminates and the maximal number of trials up to a final decision is equal to $ k cdot (r+c-1) $. The large expected sample sizes of the Sobel-Weiss-procedure [4] when the success parameters are small are avoided and the problem of a never ending sampling (using the Sobel-Weiss-procedure) when all success parameters are equal to 0 doesn't occur.
No takes yet. Share an insight, caveat, or question.
K.-H. Schriever (1978) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: