As far as I could understand, this works very well for a very specific logic puzzle problem set. Interested on future papers tho :)