Maximizing expected powers of the angle between pairs of points in projective space (Q2099823)

The authors consider, as a starting point, the conjecture from 1959 by \textit{L. Fejes Tóth} [Acta Math. Acad. Sci. Hung. 10, 13--19, I (1959; Zbl 0086.15406)] referring to the fact that the expected angle \(\arccos( \frac{x}{|x|} \cdot \frac{y}{|y|})\) between independently drawn projective points \(x\) and \(y\) equi-distribute its mass over the standard Euclidean basis. As a consequence of the validity of this conjecture it follows that the same measure maximizes the expectation of \(\arccos^{\alpha}(\frac{x}{|x|} \cdot \frac{y}{|y|})\) for any exponent \(\alpha > 1\). The discrete and continuous versions of this conjecture is analyzed in this paper in a non-empty range \(\alpha > \alpha_{\Delta^{d}} \geq 1\). Existence and finiteness of a critical threshold for the exponent \(\alpha\) in various environments, at which the maximizer changes discontinuously are obtained unless the corresponding Fejes Toth conjecture is true. The uniqueness of the resulting maximizer up to rotation is established. The paper concludes by presenting the continuous version in an \textit{Appendix A} by Dmitriy Bilyk, Alexey Glazyrin, Ryan Matzke, Josiah Park, and Oleksandr Vlasiuk, in which an effective bound for the critical threshold in the most symmetrical environments is proved.