I wrote several variations/algorithms of a program to solve sudoku puzzles (thanks to Project Euler, that someone linked to a long time ago, I quit at around 32% genius, not quite good enough at some of those mathy questions).
One of my variations did solely simple elimination, and I found from feeding it 1 star (easy) that all easy puzzles can be solved with simple elimination.
If I fed it 5 star, I didn't get a single number -- no elimination was possible.
So different rankings mean different things to different puzzlers, but I firmly believe a true hard puzzle gives you no easy numbers/low hanging fruit.
I don't claim to be the first and only person to have noticed this but I haven't seen anyone describe it just as I never saw anyone describe my "Uniqueness" strategy. Shortly after posting that method I was sent a link to some buried Times article which I'd never seen and which was similar but not as clearly described. I also initially sent off my "discovery" to a sudoku site months before that article appeared.[ Parent ]
