Voted "No, never." Fudging undermines my ability to make informed choices, and my ability to learn from and adapt around past choices. It may be a challenge to meet those requirements, while also producing a game that will make me happy. I believe that that is part of what the DM signs up for, when taking that mantle. It's part of why I've never taken it myself--I'm not yet convinced I'm up to the task. With Great Power and all that--and, within the microcosm of a D&D campaign, who has more power than the DM?
I wouldn't play 5e with a DM I didn't trust to 'fudge' as needed.
Despite recognizing that this is the 5e forum specifically, I generally tend to take questions like this as applying to any game. Would you say the same of games generally?
And, uh, what exactly does "as needed" mean, anyway? That's not really "yes," "almost never," or "never."
Come to think of it, doesn't agreeing to end an encounter early when the results are assumed constitute fudging?
I will try to provide a simple, concise definition of what *I*, at least, consider fudging:
Did you change a thing directly observed in the world, or a consequence/feature thereof, after the PCs observed it, without giving them a chance to know it changed?
If you answer "yes" to that question, then it's fudging. (The "consequence/feature" clause was added because I include abstract things, like "Diplomacy DC 15 to convince the Duke to help you," as a "thing in the world," even if they can only be observed by rolling something to see if it works.)
Agreeing to end an encounter early thus cannot be fudging, because it requires the consent of the players, and you can't give consent without being aware that you're giving consent. Otherwise it wouldn't be "agreement." It is merely a combat example of giving an automatic success on something that nominally
should be a roll, because the player's idea was
just that awesome. In this case, it would be automatically granting successful hits (and successful avoidance) for the handful of attack rolls that you want to skip, because the players have "already won"--much like a stirring, impressive speech would
already win the hearts of the royal council, making a roll superfluous.
Adding or removing HP from a monster that already "exists in play" (for those who use maps and minis, this would be "when figures/tokens hit the table") is fudging: you are making an invisible-yet-meaningful change to something under more-or-less continuous, direct observation by at least one of the PCs. Changing the DC for a skill roll is a similar, albeit more abstract, change; the world is no longer what it
is, once it "exists." Instead, existence itself is fluid and dynamic; choices can no longer be said to be good or bad based on the available information, because "the available information" may be right one second and wrong the next (or vice-versa). Changing who the unknown murderer is, when the party has already gathered good (if imperfect/incomplete) evidence of the "original" murderer's guilt is a wholly numberless form of fudging--but still fudging, because it means that the party's previously informed choices are now invalidated.
Removing, or modifying, a fight
before it breaks out is fine. Changing stats before a fight breaks out is fine (because combat itself is an ongoing process of learning the monster's stats--the being(s) therein are constantly under observation). Adding, removing, or modifying whole swathes of the world is fine--as long as the PCs wouldn't, or couldn't, have known differently. With the "tracking a murderer" example: you CAN change your mind about who the murderer is,
but the PCs need to be able to learn who the right murderer is. And all of these changes are also perfectly fine if the PCs
can find out about it before having to "face" it.
Note the bolded "can." I am NOT saying that they must be directly informed of any such changes. I am only saying that you give them sufficient opportunity to learn about it. This means there needs to be prior knowledge that (a) they
can confirm stuff, (b) even very good information
may change so confirmation is good, and (c) unless you expressly tell them otherwise, confirming their info won't cause enough delay that their plans would be ruined. These things should be "prior" knowledge because that way you aren't pausing at every remotely-restful moment to say, "Gee, it's
sure great you stole that guard duty roster!" or whatever and giving them a huge, "dramatic" wink. The onus is on the players to check, not you to
tell them to check, but subtle hints may be in order for particularly "important" changes.
In fact, I really think the "solve a murder" example is the best illustration of what I'm talking about. I see it as deeply unfair to "change" who the perpetrator is once the PCs have got their hands on good (again, not necessarily
perfect, but
genuinely good) evidence against the DM's original choice--unless there is an opportunity to learn, not about the "change," but about how their previously-good evidence was ACTUALLY faked/unreliable/etc. and that the chase is still on. Otherwise, even if it's "more awesome" to have Suspect Q instead of Suspect R be the real murderer, you've set the players up with mistaken information that they have (rather, had) every reason to think was good. And that's awful.
I like systems that don't encourage fudging - 4e D&D notably, and 5e is good too. I dislike how 3e/PF's extreme randomness & lethality to melee PCs can encourage fudging.
...uh...
*looks back at the quoted bit from Mr. Vargas*
Something is amiss with these two opinions placed next to each other, but I can't quite figure out what...
