Another blurb, today, and a late one at that. I've been working on what I used to refer to as a "mind-reading" problem, trying to get a targeting system to do what a user might unconsciously expect it to do in the most situations possible. I always find mind-reading interesting in game design (and it often applies to targeting), where you want to land either on the simplistic-and-cheap or expensive-but-usually-satisfying side of the spectrum. The thing that interests me is how developers typically start down the latter path assuming that just a few tweaks will get things to work nicely, but it almost always takes an extreme amount of work and testing to reach the level of kinesthetic happiness you need for a great game. Take any action-adventure game with melee combat, for example, and study -- just study -- what the character does to orient towards the right enemy on button press, even if it has an aiming system, and you will likely walk away impressed with the amount of mind-reading going on.