The probability of generating the symmetric group (Q5919604)

A classical conjecture of \textit{E. Netto} [The theory of substitutions and its applications to algebra. New York M. S. Bull. II. 83--106, Anzeige v. O. Bolza. (1892; JFM 24.0134.02)] on the random generation of groups is that a pair of random permutations in \(S_n\) generate \(S_n\) or \(A_n\) with a probability \(p_n\) such that \(p_n\to 1\) as \(n\to\infty\). This was proven in \textit{J. D. Dixon} [Math. Z. 110, 199--205 (1969; Zbl 0176.29901)] with an estimate \(p_n \ge 1-2(\log\log n)^{-2}\) for sufficiently large \(n\) and \(p_n\le 1-n^{-1}+{\mathcal{O}}(n^{-2})\) (with a conjecture that the upper bound is the correct probability). This conjecture, i.e. \(p_n= 1-n^{-1}+{\mathcal{O}}(n^{-2})\) was proven in \textit{L. Babai} [J. Comb. Theory, Ser. A 52, No. 1, 148--153 (1989; Zbl 0685.60012)] using the classification of finite simple groups.\par Using ideas from \textit{J.-C. Schlage-Puchta} [Combinatorica 32, No. 3, 309--323 (2012; Zbl 1299.05166)] and the Murnaghan-Nakayama rule, the authors give a classification-free proof of this result. Specifically, they prove that for any \(\varepsilon>0\) and for sufficiently large \(n\) the probability is \(1-n^{-1}+{\mathcal{O}}(n^{-2+\epsilon})\) for both \(S_n\) and \(A_n\).\par Note that \textit{A. Maróti} and \textit{M. C. Tamburini}, Arch. Math. 96, No. 2, 115--121 (2011; Zbl 1222.20003)] give explicit bounds \(1-1/n-13/n^2\le p_n\le1-1/n+2/(3n^2)\), and \textit{J. D. Dixon} [Electron. J. Comb. 12, No. 1, Research paper R56, 5 p. (2005; Zbl 1086.20045)] gives an asymptotic power series for \(p_n\) (with coefficients given in OEIS sequence A113869).