Computing Threshold Budgets in Discrete-Bidding Games (2210.02773v2)

Abstract: In a two-player zero-sum graph game, the players move a token throughout a graph to produce an infinite play, which determines the winner of the game. \emph{Bidding games} are graph games in which in each turn, an auction (bidding) determines which player moves the token: the players have budgets, and in each turn, both players simultaneously submit bids that do not exceed their available budgets, the higher bidder moves the token, and pays the bid to the lower bidder (called {\em Richman} bidding). We focus on {\em discrete}-bidding games, in which, motivated by practical applications, the granularity of the players' bids is restricted, e.g., bids must be given in cents. A central quantity in bidding games is are {\em threshold budgets}: a necessary and sufficient initial budget for winning the game. Previously, thresholds were shown to exist in parity games, but their structure was only understood for reachability games. Moreover, the previously-known algorithms have a worst-case exponential running time for both reachability and parity objectives, and output strategies that use exponential memory. We describe two algorithms for finding threshold budgets in parity discrete-bidding games. The first is a fixed-point algorithm. It reveals, for the first time, the structure of threshold budgets in parity discrete-bidding games. Based on this structure, we develop a second algorithm that shows that the problem of finding threshold budgets is in \NP and co\NP for both reachability and parity objectives. Moreover, our algorithm constructs strategies that use only linear memory.