One of the things that killed WotC D&D, in my opinion, is the emphasis on "balance". As in, every encounter should be "balanced" to the characters in the game. These metagaming considerations do make for an interesting competitive game, so I won't denigrate that consideration on that basis.
The thing is, though, that adventure games ("role-playing games") are not competitive games.
Now, now, I can hear some people out there already saying that there's been a long tradition of the adversarial DM whose every move was dedicated toward fiendish new ways to kill characters. I'd like to put that rumor to rest and note that it is bullshit (although not unadulterated bullshit, as I'll discuss in a moment), based on hyperbolic writing by some players who wanted to glorify their own exploits. "I got my character to 10th level, even though we had a DM who killed characters with a fiendish glee!" sounds a heck of a lot better than "My character got to 18th level under a Monty Haul DM." Now, when those writers published all of that, some new DMs read it and thought to themselves, "Oh, I guess that I should be playing against the PCs". And so a cycle of abuse was created.
No, what those fiendish DMs were really doing was something else. They were providing challenges to the players, not to the characters. One of the neat things about tabletop adventure games that will be a long time coming with computer versions is that the players can try anything. At all. And, if the designer didn't think of it, they can still try it and potentially succeed. This is an amazing innovation, derived from miniatures wargaming. See, back in the day, miniatures wargaming rules were, uh, a little sketchy. They were designed to cover some basic interactions, but beyond issues like casualties, movement, and morale, they didn't really cover strange situations. In a "friendly" game, the players would work out something between them that would seem to cover the situation, and in a tournament or otherwise competitive game, they'd enlist the aid of a neutral referee to the same end.
Now, this idea developed in early adventure gaming into the idea of the DM, who would design the scenario in which the players would manipulate their single figures (either explicitly, on the table, or implicitly, in imagination), and would also act as an impartial referee, ruling on the outcome of unusual actions (those outside of casualty infliction, movement, or the other issues covered under the rules). The DM would design situations that were interesting, that had no obvious solution (at times, she might even design a situation which had no predesigned solution), and so were challenging to the players. Success, in such a case, is not based entirely on the numbers on the character sheet, but on the skill of the player himself.
In a literary example (which I keep going to, since it is both iconic and therefore well-known, and also philosophically appropriate), Bilbo missed out on the xp for killing Smaug. Bard got those. But the Arkenstone was worth more xp , anyway. And even if Bilbo had worked himself up to 4th level and so hit his level limit, the true value of the Arkenstone was revealed in the way that he leveraged that item into a situation of a peacefully-coexisting dwarven keep Under the Mountain and Laketown. He made a difference in the world, and returned home with enough treasure to ensure his comfort to the end of his days. He didn't need a "balanced" encounter with Smaug. He used his ingenuity to set up a situation in which Smaug was defeated, despite the imbalance of the encounter.
What were his (theoretical) player's "victory conditions"? Not to level up, that's for sure. Not to gain untold amounts of wealth and power. They were to do what he did, I'd guess: make a permanent mark on the campaign world that the player could point to and say, "I had a major hand in that situation being as good as it is. And here is how it happened…" That's how stories come out of gaming, not by being scripted into the game, but by being descriptions of the game.