Characterizing Positionality in Games of Infinite Duration over Infinite Graphs

We study turn-based quantitative games of infinite duration opposing two antagonistic players and played over graphs. This model is widely accepted as providing the adequate framework for formalizing the synthesis question for reactive systems. This important application motivates the question of strategy complexity: which valuations (or payoff functions) admit optimal positional strategies (without memory)? Valuations for which both players have optimal positional strategies have been characterized by Gimbert and Zielonka [19] for finite graphs and by Colcombet and Niwi ński [15] for infinite graphs.

However, for reactive synthesis, existence of optimal positional strategies for the opponent (which models an antagonistic environment) is irrelevant. Despite this fact, not much is known about valuations for which the protagonist admits optimal positional strategies, regardless of the opponent. In this work, we characterize valuations which admit such strategies over infinite game graphs. Our characterization uses the vocabulary of universal graphs, which has also proved useful in understanding recent breakthrough results regarding the complexity of parity games.

More precisely, we show that a valuation admitting universal graphs which are monotone and well-ordered is positional over all game graphs, and -more surprisingly -that the converse is also true for valuations admitting neutral colors. We prove the applicability and elegance of the framework by unifying a number of known positionality results, proving new ones, and establishing closure under lexicographical products. Finally, we discuss a class of prefixindependent positional objectives which is closed under countable unions.

I want to thank Antonio Casares, Thomas Colcombet, Nathanaël Fijalkow and Pierre Vandenhove for so many insightful discussions on the topic.