iserith
Magic Wordsmith
See, that is pretty much how I play it, even I haven't stated that clearly. Obviously if you're doing something else, you're probably going to be too distracted to notice something sneaking up on you. However, if a player hasn't stated they're "watching for danger", but aren't doing something else, beyond walking and a little bit of talking, I still take Passive Perception into account for potential dangers. "Search for secret doors" would be something they would have to declare, and be considered a distraction.
We're on the same page here, though I do specifically ask the players to state they're keeping watch for danger so there are no misunderstandings. It's a safe assumption that's what they're doing if they say nothing otherwise. I'm just extra cautious, plus it opens the door for them to describe what they want to do and how which adds to the emergent story.
And we're talking about you're "UNLESS" sentence there, based on what robus proposed earlier for "bouldering". By setting a DC, the DM has automatically said "Fail" for anyone who's passive score doesn't exceed that. Your getting bludgeoned by boulders. Sorry. You had zero chance. Players would rather roll (or at least, know the DM had to roll) than just be told "you fail". That's exactly my point.
I think it's fine and in line with the rules for the DM to say "Fail" after the player's action declaration when that action has no chance of success given the circumstances. It's fine to set a DC a given player's passive check cannot beat, too, after the action declaration and consistent with previous rulings. What I was driving at is there are some DMs who are always going to rely on the dice to determine success and failure and players in such games quickly learn to ask for the roll no matter how good or bad their ideas are in context. This is discussed in the DMG under "Role of the Dice."
So you're really working back in the 3E "Take 10" concept, just as a DM call, not a player's. I don't get your reasoning of "if there is uncertainty as to the outcome", though. Swinging a sword has an uncertainty of the outcome. Should it be a passive check? It'd definitely make fights go faster.
I still disagree with library research for Int (Investigation) as a passive check. Sure the player effectively will never "roll" less than a 10. He will also never roll over a 10. Again, if the PC's passive score isn't high enough, they will never, ever find the thing they're looking for. An automatic fail.
I hesitate to call it "Take 10" despite its similarities. It's just a way of adjudicating for the DM to get at a result he or she can narrate. For me, it goes like this: Is the player's action declaration an automatic success or failure? If yes, narrate a result. If no, then is the player's action declaration a task that is performed repeatedly? If yes, then passive check to resolve. If no, then ability check to resolve.
As above, I think it's fine for a PC's passive check to fail against a given DC. It's just up to the DM to set fair and consistent DCs given the action declaration the player makes. There's also a rule in the DMG for when player have their characters commit 10 times the normal amount of time on a given task in exchange for automatic success. So in general I'm inclined to just grant success for that investment of time resource, but in some cases, a passive check is warranted. The context will tell.
With a roll, the skilled Investigator will still usually be the one to find the info, likely in less time. There's sufficient niche protection. The fighter may fail bashing the door down on a roll (my players use crowbars to get Advantage), but if the wizard then comes up and does it, the fighter can claim to have loosen it for him. Niche protection can come from narrative and role-playing too. Plus, if something requires a particular skill to accomplish, the party will generally have the person with the best score attempt it, which also re-enforces niches.
Wait, what? Are you suggesting even things like bashing a door down, which is not a repetitive actions, be handled with passive checks? So the DM could just look at players stats and skills, set a DC, and know whether each character will succeed or fail at most tasks ahead of time? That's not action resolution, that's predetermination. Really, is that what you're suggesting?
No, I am just referencing an old D&D trope for effect - unlikely character A succeeding where likely character B failed, which parallels your example. Har har, everyone has a laugh. But I think ultimately it's at character B's expense, so I can do without it. I'll seek my humor elsewhere.