Premium
Hide‐and‐seek games on a tree to which Eulerian networks are attached
Author(s) -
Alpern Steve
Publication year - 2008
Publication title -
networks
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.977
H-Index - 64
eISSN - 1097-0037
pISSN - 0028-3045
DOI - 10.1002/net.20235
Subject(s) - path (computing) , eulerian path , disjoint sets , stochastic game , mathematics , combinatorics , tree (set theory) , point (geometry) , game tree , path length , computer science , discrete mathematics , game theory , repeated game , mathematical economics , computer network , geometry , lagrangian , mathematical physics
We analyze the hide‐and‐seek game Γ ( G ) on certain networks G . The hider picks a hiding point y in G and the searcher picks a unit speed path S ( t ) in G , starting at any point S (0). The payoff in this zero‐sum game is the capture time T = T ( S , y ) = min{ t : S ( t ) = y }. Such games have been studied before, but mainly with the simplifying assumption that the searcher's starting point S (0) is specified and known to the hider. We call a network partly Eulerian if it consists of a tree (of length a and radius r ) to which a finite number of disjoint Eulerian networks (of total length b ) are attached, each at a single point. We show that for such networks, a strategy consisting equiprobably of a minimal (Chinese Postman) covering path and its reverse path is optimal for the searcher, while the optimal hider strategy is to assume that the searcher must start at the center of the tree, and to optimize in that (known) game. The value of the game Γ ( G ) is a + b /2 ‐ r . This simplifies and extends a similar result of Dagan and Gal for search games on trees. © 2008 Wiley Periodicals, Inc. NETWORKS, 2008