The thing with the alarm spell example is it sounds like the gm isn’t being hugely impartial if he is allowing the assassin to take all those measures against it
This whole debate is why I try to come up with close order action drills for all the enemies ahead of time. I make the instructions appropriate to the intelligence and abilities of the monsters. This helps me to guard against any bias.
<snip>
In the case of an alarm spell, it is almost certain to work well enough against a pack of orcs or goblins at first level. If though, they are on the run from some enemy with a lot of magical power (and at higher levels this is a lot of them), then they should expect some challenge to their alarm.
I have quoted these two posts together, because I think that the second rebuts the first - in that it shows how/why the assassin who takes extensive measures to defeat the Alarm spell needn't be
unfair at all - or at least, not in any straightforward way.
Well when it's made up, how and the reasons/criteria/principles matter.
To use my example from several posts back. If the GM creates a competent hunter at the start of the scenario. Then the hunter is tracking the players and the GM uses what they know about the hunter and how the alarm is positioned to make the call. That's very different from deciding they want to have a fight, so then they create someone who can bypasses the alarm.
One is taking the fiction into account to decide what happens, the other is deciding what they want to happen and then creating the fiction.
EDIT: The first is is why I roleplay, the second is destructive of the medium.
I think there are some potential challenges to fairness posed even in your first scenario, though. The GM uses what they know about the hunter to make the call - but does the hunter know
exactly when the Alarm spell was cast? If they guess, based on their knowledge of the PCs' travel pattern and the time of sunset and so on, how accurate is their estimate? Do they correctly identify the warded 20' cube? Do they correctly intuit the caster's selection of ringing bell vs the mental ping - because in the former case, they are going to release their own
silence effect to negate the ringing?
For me, the stuff above is not just theory-crafting. I posted the example in the OP because, a few days ago, I was reflecting on some GMing calls that I had to make decades ago when GMing Rolemaster. There were two PCs capable of casting a RM spell called Waiting Illusion - a triggered illusory effect - and they would often use the spell in a similar fashion to a D&D Alarm, to generate a bell or claxon sound if triggered; as well as the basic function of protecting against intruders, they would also use it eg if they had sneaked into a library, so that it would alert them if the librarian was coming into a nearby area.
And I would have to make decisions about whether particular NPCs or whatever would or wouldn't trigger the spell, based on my adjudication of the fiction, and in particular
the fiction of the trigger vs
the fiction of what the character in question would do. That can be
hard - especially because very often there
is no
what the character would do but at best
what the character might do.
One technique I would use was to resort to random rolls - say, assigning a percentage chance to a NPC doing this or that, and thereby triggering the spell or not. This is similar to what
@Bill Zebub has said upthread about choosing combat actions, and to what
@hawkeyefan has noted about the aliens in the Alien RPG. One way of looking at TB2e's framework, that Aetherial Premonition feeds into, is that it systematises this approach rather than requiring the GM to make the sort of ad hoc calls and rolls that RM required me to make, back in the day.