Really?
Okay, let me probe some then.
Consider the orc-with-a-greataxe example. There were three events the players could not control there:
- the orc's attack roll including confirmed crit
- the orc's damage roll
- the save vs. massive damage
To begin with, I don't accept this premise, as it relates to actual game play. Only by pulling this example out of any sort of reasonable context can you create an example in which the player has no input.
I note that some games begin with a combat, in which the players have no input. This is, IMHO, pulled out of reasonable context (although it is generally given context later), and it is incumbant upon the DM to limit lethality accordingly.
That said, I will agree that at some point
every encounter can narrow down to a point in which the player(s) no longer have meaningful choices. Once the orc's turn is up, and the axe is swinging, for example. And it is in this context that I answer the next bit.
To you, what are the acceptable boundaries for the odds that events players cannot control will lead to death or failure?
Eventually, if a character dies, the odds have narrowed down to 100%. If they did not narrow down to 100%, the character would not be dead.
This is acceptable, to me, if the DM has been fair, and the death or failure is simply a matter of odds. If you had a chance to Gather Information and you didn't, or the clues were laid out and you didn't see them. If the DM isn't
trying to kill you. (You can't beat a DM who is trying to kill you, after all.) There may, of course, always be issues of competence in designing/running encounters. But such issues have little to do with the general idea that it sucks that curing mummy rot keeps you from leaping into the next dungeon, IMHO.
It is never okay to throw a tantrum at the table. It is always okay to seek (or devise) a game more in line with what you enjoy.
Did that answer your question?
RC
EDIT: And here's a question of my own: Do you actually believe that, in a system wherein success is determined in whole or in part by random factors that failure is not an obvious potential consequence of that system?