Strongly meager sets can be quite big (Q5917569): Difference between revisions

A set \(A \subseteq 2^\omega\) is strongly meager if for every measure zero set \(N\), \(A+N \neq 2^\omega\). The dual notion is that of a strongly measure zero set: \(A \subseteq 2^\omega\) is strongly measure zero if for every meager set \(M\), \(A+M \neq 2^\omega\). This paper provides further evidence that these two similarly defined notions behave very differently. This was already well known: e.g., strongly measure zero sets form a \(\sigma\)-ideal while \textit{T. Bartoszyński} and \textit{S. Shelah} showed that under CH the collection of strongly meager sets is not even an ideal [J. Math. Log. 1, 1--34 (2001; Zbl 0980.03053)]. If \(f: 2^\omega \to 2^\omega\) is continuous, the image under \(f\) of a strongly measure zero set is strongly measure zero. In contrast, the authors prove, assuming CH, that there exists a continuous \(f: 2^\omega \to 2^\omega\) and a strongly meager set \(A\) such that \(f''A = 2^\omega\). The actual set-theoretical hypothesis needed in the proof is weaker than CH, namely the covering of the measure zero sets being the continuum. The following result, perhaps of independent interest, is established in ZFC: there exists a continuous \(f: 2^\omega \to 2^\omega\) such that for every measure zero set \(N\) the set \(\{z \in 2^\omega \mid \exists y \in 2^\omega\; f^{-1} (y) \subseteq N+z\}\) has measure zero.