If I make something up now and it turns out in a couple of weeks to have been the wrong thing to do (too strong a monster or whatever) killing the entire party because that's the world I made up is just silly, to me, though in a weird way I respect your choices, wrong as they may be in my own approach.
This is a point where the contrast between
content introduction and
action resolution becomes highly salient - the distinction that Gygax draws in his DMG when discussing wandering monster rolls.
If the monster would be too tough, you can not use it, or you can tone it down before you bring it into the game.
But there are also relevant questions about why the PCs are in combat with the monster in the first place. In classic dungeon crawling, the PCs are expected to scout the dungeon, identify targets and then return and hit them (see Gygax's advice to players in the closing pages of his PHB prior to the Appendices). If it turns out the PCs bit off more than they can chew, that's their lookout!
This is why Gygax particularly focuses on wandering monsters - because while the players can control the frequency of wandering monster rolls relative to their activities (by proper use of time, sensible precautions around noise and light, etc) the ultimate determiner of what wanderers they encounter is the GM's dice rolls. It's not all on them. Hence the need for a GM to use some judgment in the context of wanderers.
how is it fudging to validate the importance of skilled play any more negating than a random trial or series of random trials that, despite preparations, may deny them success? Doesn't that also negate the importance of skill since, based on the die roll, the skill had no actual effect?
Fudging doesn't erode the players' responsibility any more than taking the random result, which may do little to validate the good plans and tactics of the players.
If random results don't validate the players' good plans and tactics, then the game has perhaps been misdesigned as a game of skill!
Conversely, if the game is one in which luck can play a role then for some players of the game taking good or back luck on the chin is part of being a good sport.
If the game
breaks down when the random factors that it calls for are applied at face value, then I'm with Luke Crane - the designer shouldn't have put those random rolls it the game.
For instance, if the game is meant to be one in which (i) luck plays a role, and (ii) PCs survive and prosper despite period exposure to deadly violence, then why not give the
players the power to offset the consequences of bad luck rather than the GM (eg via a fate point system)? Then the players get to make the choices about what costs to bear, what consequences to offset, etc.
Frankly, I'd rather have a DM ignore a rolled failure if my plan was good rather than be stymied, after a substantial amount of prep, simply because I "rolled poorly".
If the plan is good, then why are the dice being rolled? Why do the rules call for randomness if the players don't want it and the GM is going to override it?
The solution to this problem is not GM fudging, in my view - it's getting better rules, that only call for rolls when the table is ready to accept the random outcome.
Skilled play isn't negated by fudging, it is actually facilitated by it, especially if through excellent play the PCs do everything exactly right but all end up dead because the DM rolls too many naturals 20s in a row.
In Gygax's case, he deals with this sort of situation by suggesting that rather than death the outcome be unconsciousness/maiming.
In my own games (which aren't run Gygax style, but which don't involve fudging) I handle this via fail forward (and Gygax's idea of maiming rather than death can be seen as an early attempt at implementing the same general approach). So if the dice yield a "TPK", the PCs regain consciousness in a dungeon (having been dropped unconscious rather than killed, which is an option for 0 hp in both 4e and 5e).
This another example of changing the rules to make randomness play the desired role in the game, rather than using rules that say one thing but then having the GM fudge things to make the outcome in play a different thing.
There are numerous things a DM does in a game that are for the best of the game, from encounter design to the distribution of rewards for challenges that players aren't privy to, how is overruling a die roll any different? It isn't.
As I've already posted upthread, this is dependent on playstyle.
For some playstyles - Gygaxian skilled play is an example, but not the only one - there is a big difference between introducing content into the game, and resolving declared actions. I also don't understand why you say that players aren't privy to the distribution of rewards for challenges. In plenty of D&D games the rules for distributing XP are quite transparent, and in some 4e games the rules for placing treasure likewise.
I think at some point many people forget that RPGs are really just a game of pretend with rules to make us feel less "immature."
If I as the DM can decide that my 2nd level party have stumbled on a cave with a 24th level red dragon inside that easily wipes the party with one breath attack, I am in no way playing a game nor am I being unfair.
This might be true for your approach to the game, but I don't think it generalises.
When I GM a game I am a player, bound by rules and guidelines like everyone else. And I can do things that are unfair (just as a referee can do things that are unfair), though in my own case I try to avoid doing so!
