On a game theoretic cardinality bound (Q491794)

Ever since \textit{A. V. Arkhangel'skij} [Sov. Math., Dokl. 10, 951--955 (1969); translation from Dokl. Akad. Nauk SSSR 187, 967--970 (1969; Zbl 0191.20903)] proved that Hausdorff first-count\-able Lindelöf spaces have cardinality at most \(\mathfrak{c}\) there have been attempts to weaken the various assumptions in this result. Weakening `first-countable' to `points \(G_\delta\)' does not work outright: \textit{S. Shelah} [Contemp. Math. 192, 91--101 (1996; Zbl 0847.54004)] and \textit{I. Gorelic} [Proc. Am. Math. Soc. 118, No. 2, 603--607 (1993; Zbl 0783.03028)] have provided (consistent) counterexamples. In this paper the authors improve a result by \textit{M. Scheepers} and \textit{F. D. Tall} [Fundam. Math. 210, No. 1, 1--46 (2010; Zbl 1229.54031)] and show that, assuming the Continuum Hypothesis, a space with points~\(G_\delta\) in which player~II has a winning strategy in the \(\omega_1\)-long version of the Menger game has cardinality at most~\(\mathfrak{c}\). The game has players~I and~II which choose open covers and members thereof, respectively; after \(\omega_1\) such moves player~II wins if her choices form an open cover.