Sequential Dice Game Expected Payout

This quant interview question sits at the intersection of conditional expectation and path-dependent payoffs. It takes a familiar random device and turns it into a stopping-time game where the payoff depends on how long a particular pattern persists. The setup forces you to think of the outcome as a random sum with a random horizon, a structure that appears constantly in quant finance.

It trains conditional expectation, the Law of Total Expectation, and working with recursive definitions of random processes. You must be comfortable formalizing a dynamic process, identifying a natural conditioning event, and translating a verbal game into expectations that relate to each other.

This matters for quant interviews because pricing, risk, and execution problems often have similar restart or continuation features. Being fluent with these ideas is core quant prep for top trading firms.

What it tests

This problem class is governed by the Law of Total Expectation, which allows us to decompose a complex, path-dependent random process into simpler conditional expectations based on the outcome of an initial event. The key structure is that the process has a recursive or memoryless property: after each step, the process either terminates or resets in a way that is statistically identical to the original, except possibly for a changed parameter (here, the fixed face). This recursive symmetry means that the expected value can often be expressed in terms of itself, leading to solvable equations. The principle holds because conditioning on the first event reduces the randomness of the process, and the future evolution can be analyzed as a function of this initial condition. This approach is powerful whenever the process restarts or continues with a similar structure after each step.