It seems to be common knowledge among Risk board game afficionados that certain versions have better boards than others. The original is about the best one, while the Lord of the Rings version is roundly criticised. The main point of contention seems to be "choke points" connecting a few large continents. I wonder if that can be expressed mathematically and used to evaluate potential Risk boards or even build new ones that will be recognised as "good". If so, it would be very easy to avoid publishing a terrible version.
Mokalus of Borg
PS - The quality of the board probably depends on specific rules, too.
PPS - So a board with variant rules might fare better under those rules.