Some empirical data, with each point the average of 50 trials:
<br />
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<img src="http://cs.smith.edu/~orourke/MathOverflow/MinDistSquare.jpg" alt="MinDistSquare" />
<br />
The curve $1/(\sqrt{2} n)$ (displayed) fits rather well.  Congrats to Douglas Zare and Ori Gurel-Gurevich!