System presented here can be used as an addition to classic NPC-oriented quests, but I think it would work best in a rogue-like game, as it fits nicely into randomly generated content.
The basic idea is this: in order to earn experience and advance a character, allow the players to compose their own quests.
Let’s assume that there are no NPC quests whatsoever. When a hero is interacting with NPCs, she is gathering information about what they want, desire, fear, etc. They can of course ask the hero to help them, but that doesn’t result in formal entry in the quest system log.
Instead, player has a separate “Quest Composer” screen, where he can create (though not delete) new quests, and each quest comprises goals.
Goals can be anything – ‘become king’, ‘kill Antonio’, ‘get 1000 gold’, ‘become owner of a farm’, etc. Whenever the player adds a goal, game engine computes an experience reward appropriate for goal difficulty w/r to current game state and hero statistics. Goals can have some additonal conditions attached, like ‘in 1 hour’, ‘spending not more than 500 gold’, and so on.
However, what makes it more interesting is composition. Goals can be arranged into logical trees with ‘or’ and ‘and’ nodes.
Nodes affect likeliness of quest completion – i.e. it’s harder to complete two goals connected with ‘and’ than only one, and it’s easier to complete a quest composed of two ‘or’-connected goals than both of them considered as single. Engine takes this into account when computing rewards – multiplying rewards for ‘and’ nodes, dividing for ‘or’ nodes.
Exact algorithms for computation may not be so obvious and will require extensive testing. It’s also not very obvious if it will actually be a fun to play with such a system, but I guess answer to this question requires a good prototype.
Engine should generate some goals according to the game context, i.e. information gathered from the NPCs. Solving their problems should be rewarded as in any classic RPG, but players could also add some constraints and connected goals to create more challenge for themselves and increase rewards.
I think that it can be some solution to the problem of engaging content generation. Shifting a burden of creating story to the player’s responsibilites may at first be hard, I guess – gamers are not very used to such a system – but on the other hand it introduces new dimension in gameplay and even more possibilities for meta-game (like, “let’s see who can create and complete the hardest quest!”).