THE BAD GUY WINS

Well there's a couple of approaches that I can think of off-hand. There's the multiple endings story where you can have one ending where the hero sides with the villain and they rule the galaxy as father and son. That way you can let the player's action pick the ending. (or let the player pick the ending at the very end if you're that kind of designer)

Another way is if it's you know the scope of the story. If the point of the story is to defeat the villain then obviously the villain cannot win. But if the point of the story is something else then evil may triumph just because the fact that good would win didn't even enter into it. If it is a revenge story then the character may take revenge much to the glee of whoever is next in line. If the character has to find something to save his people he might do that and maybe pave the way for something else from where-ever he found that thing.

The third way is to set up for a sequel. Let the good guy be triumphant and then the bad guy comes in and says "well, well, I manipulated you into all of this all along suck on that. *poof*" and in the next game the player is out for the head of whoever manipulated him.

Please don't ever make an "lol buy the sequel" ending. >_< If you're going to use this style of ending, do it for its own sake, for the inherent drama of the plot twist.

JustRob, I think the type of story you're talking about is a classical tragedy. In modern times, tragedies have become a very unpopular type of story. However you can find a lot of them in classical literature - several of Shakespeare's plays are tragedies for instance, as are many of the surviving stories that the Greeks wrote. Perhaps the most famous is Shakespeare's Romeo and Juliet. Typically in a tragedy, what causes the downfall of the protagonist is not that he loses the final battle, but that he has some sort of fatal character flaw. It's worth noting that these stories typically end badly for everyone - it's not like the villain usually gets a happy ending either. If the villain survives the final encounter and successfully enacts his plan, it's good to show the plan ultimately failing somehow, and possibly causing the villain's downfall as well. One of the purposes of a tragedy is to trace the path of human failure. Winning the final battle via self-sacrifice does not count as a tragedy - for a story to be tragic, the protagonist has to actually fail to achieve his goal.

From a more practical, gameplay-driven point of view, games are built around presenting challenges to the player and then requiring the player to overcome them. If the hero simply loses the final battle, giving a special ending in which the villain wins is fine... but how is this different from a game over? After watching this ending, shouldn't they be prompted to reload and try again? This is a problem you have to solve. Maybe your game can end with the protagonist defeating the villain but still failing to achieve his goal - the world still ends, his girlfriend is still killed, etc. Maybe you can make the player control the antagonist for the final battle. There are other options, but making the final battle an unwinnable battle is probably not a good way to do it. The player will feel like they simply got a game over.

The protagonist winning the day and getting everything he wants at the expense of becoming a soulless monster is still tragedy, in my book, especially if he loses the respect of the people he was trying to protect in the first place. Darth Vader is a good example of this kind of character.

Also, I am going to clear up any protagonist/antagonist confusion here. The protagonist is always the primary point of view character, regardless of whether he is "good" or "evil." It is completely possible to have a "bad guy" as the protagonist. Similarly, the antagonist is always the opponent of the protagonist, even if the antagonist is himself "good."

There's a specific literary term for this type of character, a tragic hero whose flaws or obsessions ultimately end up being his downfall. It's called the "Byronic Hero" and if you're interested you may want to do a bit of research about it.

If you make an epic game with likable characters and in the end they lose to the evil dude even AFTER a highly motivational speech, be prepared to have hit and miss feedback from others.

If there is both a bad and good ending and both sides are awesome then I would not mind it one bit.

In my game, I am very carefully going out of my way to ensure that the antagonists aren't presented as moustache-twirling, cloak-flipping villains who appear briefly to berate you and to summon boss monsters. The ultimate goal/s that the two sides represent should be more of a moral conundrum than a matter of flat good and evil; I guess, in short, I'm subverting this - by trying to make the bad guys not look so bad. An additional layer is that the world setting is pretty bleak; it's not going to make much difference if you "SAVE TEH WORLD" if the world isn't very worth saving to begin with.

As far as "lol buy the sequel" - I'd say don't do it just for that effect, BUT, I do love endings that leave some things open to interpretation.

This could work, and set up precedent for future games without leaving so much to be desired that players are forced to buy subsequent games just to get the whole story.

For example.

Hero A, lets call him Mark, is not really so much a hero. One of those reluctant heroes you hear so much about. He is dragged into a large conflict because his Fiancée, lets call her Kelly, is taken along with all the other villagers in his hometown for some sinister plot by the main villain. Mark manages to avoid it, somehow, and sets out to rescue his Fiancée.

Done properly, this could be the ultimate goal for your hero. Save his girlfriend. That doesn't necessarily mean beat the bad guy, or save the world, or even rescue the other villagers. It would take a strong narrative muscle to manage to progress the villain's story at the same time so that the game doesn't appear to just be about Mark and Kelly, with all sorts of other stuff going on that is mostly meaningless, but it could work and would offer something slightly off the beaten path.

What I meant though, was that you aren't tasked with thwarting armies and finding magical artifacts for other people, just to find your way to a small goal. The point was that that small goal was actually a huge goal in the eyes of the Hero, which he is more likely to be enlisting others to help him do, as opposed to helping others on his way there. Like I said, it can be done. But it would require very strong narrative skills.

While a few of us are explaining what we seek to accomplish with our games that lie just outside of "good hero saves world. END", you could make the world "un-save-able" because the villain is killed, either out of a balance or him being the only one who knows how to restore the decaying land. You could also have a villain that led a life of evil called for by a "master" figure, but strives towards redemption when he realizes only too late that his life went to shit because of his choice, even if not in the hero's eyes but in the eyes of the governing deities. This way he doesn't necessarily fall into the "bad guy turned good" character, but more of a villain who achieved his own end, which in my game feeds into what I was explaining earlier.
Somehow, I swear I'll make the world coming to an end work.
If your game just so happens to follow the cookie-cutter, completely uninspiring "good hero save world. END" storyline, any greater attempt at having the bad guy win will seem nothing more than a slap in the face after all that the player went out of their way to do.