Could you point me to the heuristics that say that violence is always a bad strategy?

There's nothing to point at; that's the whole heuristic. While there are lots of good arguments for violence being bad in general and in specifics, the only thing a heuristic requires is that when asking "is violence a good strategy here?" the answer is almost always "nope". Which it is, but for a slightly different mix of reasons every time you ask it.