The Adequacy of Universal Strategies in Analytic Gambling Problems

Suppose measurability structures are imposed upon a Dubins and Savage gambling problem. A long-standing question asks whether strategies which are measurable with respect to these structures enable the gambler to maximize his return. Measurable strategies have the advantage over arbitrary strategies in that they induce countably additive probability measures. In this work, we show that measurable strategies are adequate in order to maximize return if and only if the optimal return function is measurable. Using this result, several examples of gambling problems for which measurable strategies are adequate are given.