Unraveling \(\boldsymbol\Pi_1^1\) sets (Q1591205)

Martin's inductive proof of Borel Determinacy introduced the concepts of covering and unraveling. In this paper the author constructs a covering that unravels \(\boldsymbol\Pi^1_1\) sets. This requires a large cardinal assumption. Work of \textit{J. R. Steel} [``Determinacy in Mitchell models'', Ann. Math. Logic 22, 109-125 (1982; Zbl 0573.03026)] shows for the relevant large cardinal \(\kappa\), \(o(\kappa)=\kappa^{++}\), where \(o\) is the Mitchell order on measures. The author uses a cardinal \(\kappa\) such that for every \(Z\subset V_{\kappa +1}\) there exists a normal, \(\kappa\)-complete measure \(\mu\) on \(\kappa\) so that \(Z\) belongs to Ult\((V,\mu)\). For his \(\kappa\), \(o(\kappa)=\kappa^{++}\). Using such a cardinal, he shows how to unravel a single \(\boldsymbol\Pi_1^1\) set and then a countable collection of \(\boldsymbol\Pi^1_1\) sets. Using these results, the author shows the determinacy of games with a \(\mathcal A \mathbf\Pi_1^1\) payoff (\(\mathcal A\) is the Suslin operation). Finally, the author shows how to unravel sets in the \(\sigma\)-algebra generated by \(\boldsymbol\Pi_1^1\) sets, obtaining determinacy for games with such payoff sets. He explains why this result is optimal.