Meager-nowhere dense games. I: \(n\)-tactics (Q1209562)

Banach-Mazur game for a topological space \(X\) is played as follows. Player I picks a nonempty open subset \(U_ 1\) of \(X\) and then player II picks a nonempty open subset \(U_ 2\subseteq U_ 1\), and then player I picks a nonempty open subset \(U_ 3\subseteq U_ 2\), etc., etc. This play of the game is won by player II iff the intersection of the \(U_ i\)'s is nonempty, otherwise it is won by player I. This game was invented by Mazur in the 1920's, who conjectured that for \(X\) a subspace of the reals player II has a winning strategy iff \(X\) is comeager. This was proved by Banach. The games of this paper were suggested by considering the question of tactics or \(n\)-tactics in the Banach-Mazur games. An \(n\)-tactic is a strategy in an infinite two person game which only depends on knowing the last \(n\) or fewer moves of the opposing player. Part I of the paper is concerned with the game \(MG(J)\), the monotonic meager-nowhere dense game. This is played as follows. Given an ideal \(J\) on \(X\), let \(\langle J\rangle\) be the \(\sigma\)-ideal generated by \(J\). The players successively choose \(M_ 1\), \(N_ 1\), \(M_ 2\), \(N_ 2,\dots\) such that \(M_ k\in\langle J\rangle\) are chosen by the first player with \(M_ k\) a proper subset of \(M_{k+1}\) and \(N_ k\in J\) are chosen by the second player. The second player wins this play of the game \(MG(J)\) iff \(\bigcup_{k<\omega}M_ k\) is contained in \(\bigcup_{k<\omega}N_ k\). One of the results obtained is that the second player has a winning 1-tactic in \(MG(J)\) iff \(J=\langle J\rangle\). It is also shown that if the continuum hypothesis is true and the cofinality of \(\langle J\rangle\) is at most \(\aleph_ 1\), then the second player has a winning 3-tactic in \(MG(J)\). Many other results are proved about \(MG(J)\). In Part 2 the game \(SMG(J)\) is defined. It is the same as \(MG(J)\) except the first player must choose \(M_{k+1}\in\langle J\rangle\) such that \(M_{k+1}\) contains \(M_ k\cup N_ k\). One of the results proved is that if \(\langle J\rangle\) has cofinality \(\aleph_ 1\), then the second player has a winning 2-tactic in \(SMG(J)\).