Colorability Saturation Games (1606.09099v6)

Abstract: We consider the following two-player game: Maxi and Mini start with the empty graph on $n$ vertices and take turns, always adding one additional edge to the graph such that the chromatic number is at most $k$, where $k \in \mathbb{N}$ is a given parameter. The game is over when the graph is saturated and no further edge can be inserted. Maxi wants to maximize the length of the game while Mini wants to minimize it. The score $s(n,\chi_{>k})$ denotes the number of edges in the final graph, given that both players followed an optimal strategy. This colorability game belongs to the family of \emph{saturation games} that are known to provide beautiful and challenging problems despite being defined via simple combinatorial rules. The analysis of colorability saturation games has been initiated recently by Hefetz, Krivelevich, Naor, and Stojakovi\'c (2016). In this paper, we improve their results by providing almost matching lower and upper bounds on the score of the game that hold for arbitrary choices of $k$ and $n>k$. In addition, we study the specific game with $k=4$ in more details and prove that its score is $n^2/3+O(n)$.