Expected Dice Game Payout

This quant interview question is about understanding a simple stochastic game with a stopping rule and a running payoff, framed in terms of conditional expectation. It appears deceptively easy, which is exactly why it is popular in quant prep: it reveals very quickly whether a candidate can translate a verbal description of a random process into a clean mathematical object. You need to see beyond the die to the underlying expectation structure.

It trains the ability to formalize recursive expectations, apply the law of total expectation, and work comfortably with infinite horizons in a controlled, discrete setting. It also sharpens intuition for when a process "restarts" and how to exploit that to express an unknown expectation in terms of itself. This is core to modeling, risk, and pricing problems in quant interviews.

This matters for quant interviews because so many trading and risk processes have similar restart or continuation features: you observe something, sometimes stop, sometimes continue, and care about the expected payoff. Interviewers use questions like this to probe whether you can identify the state, structure the randomness properly, and compute the fair value without getting lost in the mechanics. For quant prep, mastering this pattern pays off across many seemingly different interview questions.

What it tests

Problems of this class are governed by the principle of recursive expectation, where the expected value of a process is expressed in terms of itself, conditioned on the possible outcomes of the first step. This is a manifestation of the Law of Total Expectation, which allows you to decompose a complex, potentially infinite process into manageable cases based on the immediate next event. The key is recognizing that after certain outcomes, the process "resets" and the expected value from that point onward is the same as the original expectation. This recursive structure holds because the process is memoryless after each round: the future is independent of the past, given the present state. The pattern persists in any process where the continuation depends only on the current outcome, not on the entire history.