MathOverflow will be down for maintenance for approximately 3 hours, starting Monday evening (06/24/2013) at approximately 9:00 PM Eastern time (UTC-4).

For i.i.d. points chosen in a bounded subset of $\mathbb{R}^n$ \mathbb{R}^3$(or$\mathbb{R}^d$) it seems to me that$\theta_\min(n)\to 0$is ensured when the support of the distribution has a smooth boundary. This covers the case of the uniform distribution on an Euclidean ball, and a uniform spherical distribution as well. (I'm not quite sure about how to state a converse). 2 added 22 characters in body For i.i.d. points chosen in a bounded subset of$\mathbb{R}^n$it seems to me that$\theta_\min(n)\to 0$is equivalent to: the convex hull of ensured when the support of the distribution has a smooth boundary(I'll add details at request). This covers the case of the uniform distribution on an Euclidean ball, and the a uniform surface spherical distribution as well. (I'm not quite sure about how to state a converse). 1 For i.i.d. points chosen in a bounded subset of$\mathbb{R}^n$it seems to me that$\theta_\min(n)\to 0\$ is equivalent to: the convex hull of the support of the distribution has a smooth boundary (I'll add details at request). This covers the case of the uniform distribution and the uniform surface distribution as well.