You've already described what the rules say to do. That's literally the rules describing the processes of play.
And that's before we even get to Rule Zero, which I know you love... which basically says the DM can do whatever they want.
In that sense, you could argue that all of D&D is contained in one rule.
The rule of DM fiat.
Referees in football and other sports have to make judgement calls all the time.
Not really, no. Their judgment is in, "Did I see it hit the line or not?" or "Did I see holding or not?" They have no judgment to alter the effects of the holding call or what out of bounds means.
No, I said it absolutely is a case of changing or removing a rule.
I misread your response. However, it's not unfair.
So... if it's okay by the rules, then it's fair... but if it's not okay by the rules, then it's unfair.
This is what I said.
But it's not what I said. I've said repeatedly that some rules involve fairness. Others do not.
I mean, according to
@bloodtide it would be an issue 50% of the time.
So what. I'm not him. It's not going to be an issue anywhere near 50% of the time. I and the groups I've played in and DM'd have been using alarm since it came out and I can't remember it ever being an issue. It might have been, but nowhere near 50% of the time. I doubt even 1%.
And although this scenario is unlikely to be common at most tables... once you start adding up all the places where GM fiat leaves things so open, it becomes more and more of a concern. That was, I think, part of the point of
@pemerton 's OP... to point out how much is actually going on in a specific example involving one spell.
DM fiat is not a concern in general. And all the added up "concern" is still less than 1% of the time, so even this specific instance the DM fiat is just not a concern. If you worry about super long odds, you'll never enjoy the game, or life for that matter.
You should worry about what's likely. Not what isn't.
But in the absence of the alarm, couldn't the DM rely on Stealth rolls and Perception rolls to see if anyone wakes up? And let's say the rolls go poorly for the players, couldn't the DM decide that, with the outcome of the action not really in doubt, that the intruder is able to simply kill a PC by slitting their throat?
Sure. But in the presence of alarm, it's also not in doubt. You wake up and are aware of the threat.
As for the bolded, I've played in some games like that. But we were made aware of that possibility before playing the game. It's not something that is sprung on people.
The alarm spell does not say it prevents surprise. That's your interpretation. It's exactly the kind of "unseen" GM fiat being described in the OP.
it doesn't have to say it, because it simply does it. It says it wakes you up when an intruder shows up. You wake up and are aware of a threat.
You need to take off the blinders and stop looking just at the alarm spell. There are other rules that interact with the spell that determine who is surprised and how, and the alarm spell interacts with those rules in such a way that you are not surprised. You can only be surprised if you are unaware of the threat, which is impossible when the alarm spell wakes you up, instantly making you aware.
Again, that's your interpretation of the rules. But others may interpret things differently.
Yep. I had people arguing with me despite the rule saying this, "You automatically revert if you fall unconscious, drop to 0 hit points, or die." that the druid never hit 0 hit points through damage.
Misinterpretations are interpretations, but are not correct.