A Booby Trap Game

TL;DR

This paper models a strategic game involving a defender placing booby traps and an attacker selecting search areas, providing solutions for various geometric and network configurations including Euclidean spaces and specific graph structures.

Contribution

It introduces a novel booby trap game framework and offers analytical solutions for different search space geometries and network topologies.

Findings

Optimal strategies are derived for Euclidean spaces.

Solutions are provided for circular and linear networks.

Bounds are established for tree networks with multiple traps.

Abstract

This paper presents a booby trap game played between a defender and an attacker on a search space, which may be a compact subset of Euclidean space or a network. The defender has several booby traps and chooses where to plant them. The attacker, aware of the presence of these booby traps but not their locations, chooses a subset of the space and collects a reward equal to the measure of the subset. If the attacker does not encounter any booby traps, then the attacker keeps the reward; otherwise, the attacker gets nothing. The attacker's objective is to maximize the expected reward, while the defender's objective is to minimize it. We solve this game in the case that the search space is a compact subset of Euclidean space, and then turn our attention to the case where the search space is a network in which the attacker must choose a connected subset of the network. We solve the game when the network is a circle or a line. For the case of one booby trap, we solve the game for 2-connected networks, and when the network is a tree we present an upper bound and a lower bound for the value of the game whose ratio is at most 27/25. We also present an optimal solution for each player in a few cases where the tree is a star network.