Same. I only change something after it has been encountered if it's broken somehow, and then I let the player know why I'm changing it.
If you're telling the player(s), it's not fudging.
As I said, I only do it for extreme bad luck and not just bad luck.
Max, this is exactly what makes it fuzzy. Who defines "extreme" bad luck? Why is three failed death saves after a failed opening banshee save not "extreme bad luck" but (say) three crits in a row is "extreme bad luck"? That is the very fuzziness you claim to have removed.
How would you feel if the DM uses fudging (or techniques akin to fudging) to instead prevent an outcome that goes against the expectations and understanding of the players?
I'm not sure I understand what you mean. But, at least for me, unexpected results are not inherently a problem, and miscommunication is better solved by good communication, not by secret under-the-hood corrections. Especially because if there's a misunderstanding, the DM may not understand why it happened, thus creating the possibility of a cascade of changes until the DM's side of the misunderstanding is resolved.
Also, does your answer change depending on the reason an outcome that goes against the expectations and understanding of the players was possible in the first place?
I don't think so?
(I'd see correcting the misunderstanding as preferable, but potentially impractical depending on how long ago the decisions based on that misunderstanding were made.)
I would prefer the impracticalities of correcting a long-held player misunderstanding, personally, regardless of where I sit at the table. As DM, I occasionally fail to clearly communicate to my players, so I find the process of correcting these issues openly to be better for both me as someone refining my DM skills and for my players as people learning how to play the game.
Thanks for the detailed reply! I agree that correcting the miscommunication is the preferable response when it's quick and easy. If fixing the miscommunication involves winding back past decisions (as can be the case when the decision was a strategic one, rather than a tactical one), however, I think fixing the miscommunication behind the scenes (whether via fudging, something akin to fudging, or some other method) becomes a useful alternative for validating the players' decisions.
I still think that invalidates the decision-making. It just does not do so by taking away the benefits they have earned from doing so.
It is the equivalent of secretly turning on a handicap in a game so that your SO feels good about choosing a difficult character in a video game when they misunderstood you and thought you were saying that character would be easy to learn. (Consider something like Pokken, if your SO wants to play solo because they like Pokémon and want to learn more about the fighting games you enjoy without all the competition, so they're playing it solo and you help set it up.) Their choice has no actual relation to the consequences that occurred, you as facilitator made those results possible.
So, not fudging in the dice sense (although it has probably greatly lessened the cases I've been tempted) - but how does it feel if the bad guys need to confirm their criticals, and have a chance of fumbling if confirmed, but the players just use the normal rules? (I was thinking of doing it for the players too, but forgot to tell them until we were well into playing, and it's just stuck...)
This is harder because the three systems I favor most (DW, 4e, 13A) don't use crit confirm rules. But if this is a consistent rules difference...I guess that is fine? It would be a bit dubious if the players aren't allowed to know about this difference, but it wouldn't upset me greatly. More like "oh, I kinda wish you'd told us that a while ago!"
For comparison, there's a vaguely similar thing in 13A, the Escalation Die. You have probably heard about it but just in case: after every completed round where the action advanced (hence "escalation"), the DM either puts the Escalation Die (d6) on the table with the 1-pip face up, or advances the die to its next face, hitting a maximum of 6 at the start of the seventh round. Player characters add this value to all attack rolls and various other things. Monsters do not....except dragons, because dragons are SCARY and should not be taken lightly. This is an official, player-facing rule, made clear from the outset. Further, while it might
sound like it's blatantly pro-player, what it actually does is enhance the valid strategy space by discouraging the usually hyper-dominanr "nova" tactic: hit percents are tweaked down very slightly (1-2 points below the usual "expected" 60%-65% hit rate), so that there's an incentive to hold onto your big guns until you are confident they will hit, balanced against the potential reward for using them early and getting lucky. So, while it IS beneficial to players in the long run, it is actually a (mild) detriment initially, which helps address a long-standing unfortunate (and boring) pattern in D&D strategic choices.