I also appreciate Retina's symmetry argument:
put a dart in each region, and count the number of ways
of removing them to reduce the total to the required number.
He was relying on statistics to guess that the maximum wasn't
located in twin peaks equidistant from the middle.

David

[QUOTE=davieddy;125754]He was relying on statistics to guess that the maximum wasn't
located in twin peaks equidistant from the middle.[/QUOTE]Yeah I was, it is quite a safe bet with the distribution given, and also the fact that I could easily verify the fact by examining the output of the coefficient list.