Random squares is absolutely not cheating -- if you really really want a grid, you can always draw one out from the square you found. Indeed, not picking the g

Random squares is absolutely not cheating -- if you really really want a grid, you can always draw one out from the square you found. Indeed, not picking the grid that gives you the best answer might be considered cheating :). Unless you're explicitly going for OS grid reference square kilometres, which the first grid may have been -- but I see no evidence that the author did so. And if you want to really not cheat, you probably need to consider rotating your grid too. All that said, I concur that it probably doesn't change the result very much. I'm just slightly disappointed that the author thought that improving his method to get a more correct answer was "cheating".