Rigorous Computer Analysis of the Chow–Robbins Game
From MaRDI portal
(Redirected from Publication:5404048)
Abstract: Flip a coin repeatedly, and stop whenever you want. Your payoff is the proportion of heads, and you wish to maximize this payoff in expectation. This so-called Chow-Robbins game is amenable to computer analysis, but while simple-minded number crunching can show that it is best to continue in a given position, establishing rigorously that stopping is optimal seems at first sight to require "backward induction from infinity". We establish a simple upper bound on the expected payoff in a given position, allowing efficient and rigorous computer analysis of positions early in the game. In particular we confirm that with 5 heads and 3 tails, stopping is optimal.
Recommendations
- A multi-computational exploration of some games of pure chance
- Exact algorithms for solving stochastic games
- scientific article; zbMATH DE number 4140924
- Algorithmic rationality: game theory with costly computation
- Numerical analysis of ruin games
- The complexity of stochastic games
- Automata, Languages and Programming
Cited in
(6)- On the \(\frac{S_n}{n}\) problem
- Computer solution to the game of pure strategy
- scientific article; zbMATH DE number 687938 (Why is no real title available?)
- scientific article; zbMATH DE number 4184524 (Why is no real title available?)
- An experimental mathematics perspective on the old, and still open, question of when to stop?
- On the minimum expected duration of a coin tossing game
This page was built for publication: Rigorous Computer Analysis of the Chow–Robbins Game
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5404048)
