No matter how interesting a mythology is, there are several problems with it being used in a game. The game will be benefiting from copying the mythology with no real adherence to it; and the game has certain other goals that dictate how a scenario will play out and what the players' options in it are. Someone you are determined to kill will interest you only so much, and may interest other players even less. No matter how skillfully the story and graphics are presented, the game may contain elements that cause the average person to reject this presentation due to the implicit support of these motives or values. A similar situation may also arise for lore that the game creates itself, as people often believe in a thing only because they feel its base elements benefits themselves or others and no story is real enough to stand without belief.
Things like PvP ranks, and changing power scaling to reduce competitiveness would help. Just a reminder, the Ninja Law of 1 vs many helps to balance lack of power scaling, so the significance of a single low-level opponent does not translate into being destroyed when fighting multiple of them.
However, there are deficiencies in the story-based progression incentives too. When a discussion is oriented around 'strategies' to defeat a character who is significant in lore, it detracts from the story and disrespects the character who may come to be important only because they drop items upon their death. What if you did not feel you had to kill everyone you encountered?
This is a three-part balancing act: if you kill everything, then everything is only for killing. If you have too much choice in not killing, then the world becomes peaceful and the complementary option is still removed. There must be just enough choice to raise this option as a possibility and maintain it, without letting things become too predictable either way.
Stopping ongoing combat could be bad.
Discussion leading to combat may work, if it may also lead to dialogue.
The 'rewards' of a situation must not force either option. Needs of people in a group dictate the group actions, and if someone needs certain 'drops' this could force a situation. Diversity in rewards helps, so both conflict and dialogue can be acceptable options.
Short-term gain and long-term gain must be balanced with inconvenience, for determining these choices. Sneaking past a watcher. . .
Combat may lead to other results other than death, if these results are also beneficial and do not make the situation seem contrived or force themselves on the players unwanted. Specifically thinking of a certain underhanded individual at a tavern near a marsh . . .
Many scenarios become possible when a story has time to evolve. An instanced area in a game centered around story, instead of progression, can provide this time for a group to explore a story, with variations of instance binding providing the continuity. A key concept, and an expansion of the 'instance as fortress' perspective, is multiple factions within an instance. A group that makes the proper decisions may be able to actually ally themselves with one faction, allowing them to complete goals that involve conflict with another faction. When the story is done a different time, the group may come to ally with a different faction, or even go to war with both. The accumulated 'reputations' with factions of individual players may have an effect even if the reputation of a single player does not enable or deny certain options by itself, making it easier or more difficult to accomplish coercive goals. This brings into question the idea of conflict with any specific character, when this conflict may be avoided by the correct actions earlier in this enactment of the story. The right contributions to a leader who is suspicious of you could alleviate their concerns and gain a powerful ally in your fight with your opponents in obtaining your true goal . . .
There are other details in accomplishing this set of options in a game. When there is only a single standard of success and a single linear goal, everything else becomes subordinate to this goal. This is seen in current games. When multiple diverse goals offer many and sometimes conflicting standards of success, the options are not so simple. This is part of having story, and quests, be a way to differentiate and express instead of just a progression in power. Complexity of accomplishing a given task is entirely different from being able to ask whether to accomplish this task at all, while still satisfying other in-game requirements and the player's own standards of achievement and consistency of purpose. Complicated fights do not make up for lack of choice.
This means that, due to the quests they choose to accept (or other methods of differentiating character and their goals), some people may want to ally with one faction, while others will want to ally with another. It is imperative that this does not cause groups that agree ahead of time to choose only one or another path!! There should not be 'LFM
It may also be possible to incorporate the 'immunization-type' raid progression enhancements into the faction format . . . but I don't know, it's a separate idea whether or not it might be required for them to work together. For example, maybe you could only win over a certain ally if the player negotiating with them had consistently shown their rapport with them during past gameplay for several weeks . . . somewhat modified version of the faction-insensitive and combat-oriented 'immunization' system for raid progress. Exploring the complexity of story paths could take as long as actually implementing a known plan; for example, you could be required to work to befriend a faction at one point of time, but then later on need to find separate allies to betray your first ones later on in the same instance . . .
In all cases, relations should be sensitive to specific actions. Destroying a random patrol and killing all its constituent entities does not make you look good to the command that sent them out, and the attitudes of different parties in dialogue should reflect that.
Omg, two ideas down ("factions", "gain, split") and only two more to go!!1 ("death", "simple")
When a player dies, it is a very negative thing. Sometimes this is good, such as when that death could have been avoided. Sometimes this is bad, such as when the player came up against a stupidly powerful enemy and their death had no purpose, or when they died due to random uncontrollable factors in an otherwise necessary situation, or when those who were supposed to be helping them ended up being stupid or incompetent or both and there's no way to accomplish their goal and no point in being there at all, or when any of these situations leads to any other like when a stupid decision leads to an uncontrollable death which leads to a wipe due to game mechanics not allowing otherwise. Part of this is bad design that causes slight but critical mistakes to overcome the skill delta that is restrained by the gear/progress delta, but part of it is also just stupid teammates who didn't help you and didn't heal. The critical aspect is that difficulty is enjoyable when it allows you to enjoy a situation and succeed; but it is not enjoyable when deficiencies in those around you cause you to fail again and again with no benefit.
Therefore, it may help to give a slight gain for dying in the right situation.
But not enough to make you do it on purpose.
How does this relate to the other stuffs talked about above? What would be the right consequences for dying? :P I haven't thought up any details.
And finally, 'simple'!! If being with other people is fun and you are able to roleplay and enjoy a story, it does not matter if a fight has new and complicated mechanics that require specific strategies. More complex means more challenge to adapt to and memorize and overcome, but this is bad as well as good because these challenges will usually have no relevance to the real world, and it just becomes a needless complication. The original game that has become so important in this genre never had complicated fights when it started out and did better than ok . . . people look for things they can't always find. That's from a PvP movie by the way, sort of.