Topological games defined by ultrafilters. (Q1426512): Difference between revisions

Let \(X\) be a~space, \(x\in X\), and let \(p\in\omega^*\) be a~free ultrafilter on~\(\omega\). Let \(\mathcal G_p(x,X)\) be the infinite game for players I and~II in which on the \(n\)-th step, player~I chooses on open neighbourhood~\(U_n\) of~\(x\) and player~II responds by choosing \(x_n\in U_n\). Player~I wins if \(\{x_n:n\in\omega\}\) has a~\(p\)-limit point in~\(X\). A~space is said to be a~\(\mathcal G_p\)-space if player~I has a~winning strategy in \(\mathcal G_p(x,X)\) for every \(x\in X\). For \(p\in\omega^*\) let \(T(p)=\{\hat f(p):f\)~is a~permutation on~\(\omega\}\) and \(R(p)=\{\hat f(p):f\in\omega^\omega\) and \(f\)~is strictly increasing on some \(A\in p\}\) where \(\hat f\)~denotes the Stone-Čech extension of~\(f\). The authors prove that for any free ultrafilters \(p\),~\(q\) the space \(\beta\omega\setminus\{r:r\leq_{\text{RK}}p\}\) is a~\(\mathcal G_q\)-space if and only if \(q\nleq_{\text{RK}}p\); and for every \(p\in\omega^*\) there is a~\(\mathcal G_p\)-space that is not a~\(\mathcal G_q\)-space for every \(q\in T(p)\setminus R(p)\). As a~consequence, \(p\in\omega^*\) is a~\(Q\)-point if and only if every \(\mathcal G_p\)-space is a~\(\mathcal G_q\)-space for every \(q\in T(p)\).