Groups equal to a product of three conjugate subgroups (Q502976)

It is well-known that if \(G\) is a group and if \(A\) is a proper subgroup of \(G\), then \(AA^x\neq G\) for every \(x\in G\). It is then a natural question to ask if \(G\) can be written as a product of more than two conjugates of the same proper subgroup \(A\) and, if so, to find the minimum possible number of factors in the decomposition. Obviously, if \(G\) is a finite and nilpotent group such a decomposition is not possible since all the maximal subgroups are normal in \(G\). The authors employ very interesting and variate techniques to derive general information about a three factor decomposition of a general (not necessarily finite) group \(G\) as \(G=AA^xA^y\). It is shown that for \(x\in G\) the equality \(G=AA^x{A}^{x^2}\) happens if and only if \(G=(AxA)(AxA)\). After this, they unleash the power of ``computer algebra'' in order to check the relevant cases when \(G\) is a finite simple nonabelian group. The fruit of their combined efforts is the beautiful and general Theorem 2: If \(G\) is a finite non-solvable group, then there exists a proper subgroup \(A\) of \(G\) and some \(x\in G\) such that \(G=AA^x{A}^{x^2}\). The appendix at the end of the paper provides, for each sporadic simple group, an example of a subgroup \(A\) which realizes the required decomposition.