We've talked about plot, and about the promises we make our readers when we publish a story. We've talked about magic and power balance, and how they contribute to the tension that keeps a story going. So now it's time for the end of the story.
Once again, for the sake of simplicity, I'm going to call the major teams of a fantasy novel Team Good, lead by the Hero and Team Evil, lead by the Villain. I recognize that fantasy protagonists and antagonists come in a lot of different shades (Next Post: The Cast) but these terms will do the job for the sake of today's discussion.
Here's the claim that might get a lot of comments: in a straight fantasy, Team Good always wins. The readers do not want to spend however many hundreds, if not thousands, of pages watching Team Good get to the point where it can defeat Team Evil and then lose. Not in the final battle. If Return of the Jedi had ended with the Rebel Alliance being annihilated, and Darth tossing Luke off the bridge while the Emperor cackled in the background, fans would have rioted and Lucas would have never made another movie.
Now for the huge mess of caveats:
Team Good must win is not the same thing as the Hero has to come out alive, or that there is any guarantee of a happy ending for any individual member of Team Good. As long as Team Good wins, or the actions of Team Good assures a future win, you can kill every single member of Team Good and it's still a satisfying end. If you re-wrote the movie Glory and set it with a troop of Dwarves looking to be taken seriously as real members of Elven society, the ending where almost everyone dies, but the first spark of Dwarven equality bursts to light, then the ending works. Likewise, if your Hero comes out of the final battle wounded, disillusioned, broken, having lost everything but his ability to breathe, as long as whatever he was trying to accomplish gets done, all is good.
Team Good must win only applies for the final conflict. You can write a fifteen book series and have Team Evil happily torturing the Hero at the end of the first fourteen of them as long as the Hero triumphs at the end of the last book. Romances demand a Happily Ever After for each book in the series. Fantasy demands that Team Good has to win in the end.
Winning is entirely dependent on how you defined the problem. If the issue is Dark Lordenstein is big, bad, and EVIL, any end where Dark is no longer those things flies. Team Good can kill, convert, imprison, or do any other thing you can imagine to Dark, as long as Dark is out of the picture as the Villain in the end.
You'll probably notice that I'm dealing mostly with the man v. man version of fantasy. Now, while it is true that man v. nature and man v. society do happen in fantasy, they are almost always secondary plot lines or some level of meta problem that's being explored in a micro man v. man setting. But, if you've got it set where your protagonist is fighting all of society, or magic, or whatever external, non-person (using person loosely here, by this I mean a thinking creature, not necessarily human) problem he or his team still has to triumph in the end.
If the problem is societal, like racism, winning can be defined as any step that causes the society at large to seriously consider change, but does not require immediate change. If the fight is the main character versus his nature (he's a werewolf or vampire or whatever) whichever side you've set as the protagonist needs to take the day.
If you've written the sort of epic where there are seventeen main characters, no clear protagonist, and all of them are dingy shades of gray without a Team Good or Team Evil, more power to you, 'cause that's a bugger to write a satisfying ending to. No matter how you end that, someone is going to be pissed because his or her pet character didn't get the ending they were looking for.
Basically, the more clearly defined the problem is, the easier it is to write a good (once again, good is defined as the ability to satisfy the majority of readers) ending.
Okay, so here's the one big exception to this rule: tragedy is an acceptable version of fantasy. There are a few main variations on this where Team Good doesn't win and the readers don't mind.
Tragic Fall: call this Macbeth style fantasy. You're writing about the fall of someone who did something stupid. Usually in this case the Hero is very dark gray if not outright black, but he's still a compelling POV character. These are often morality tales about the love of power and how it destroys people. You can do any horrible thing you want to Hero, and the readers will appreciate it.
The Sacrifice: This is the story where the chance of Team Good winning was one in a million at the beginning of the tale, and it's still one in a million at the end. Call this Les Miserable style fantasy. Everyone knows from page one that this is not going to end well for Team Good, but that they're doing something that will get them all killed because it's important. The story here is about the struggle, character development, and then breaking the reader's hearts by crushing those characters. Usually this ends with the goals of Team Good getting more attention and the possibility of a future win for Team Good, but the immediate Team Good has failed and is dead.
The Anti-Happily Ever After: Romeo and Juliet style star-crossed lovers flies in fantasy as well. But this is usually a microplot. Team Good usually wins whatever it is that's going on in the metaplot, but doing so requires the star-crossed lovers to put each other aside/die. This is tragedy for whichever characters, but the greater Team Good still wins the day. If Romeo and Juliet both die, followed by the Montagues and Capulets immediately going to war and annihilating each other, your readers will feel cheated. The micro tragedy has to have a bigger meaning and value than the immediate characters.
Having read all that, you know what to do, right? Exactly, do not write the story where everything is going swimmingly, the odds are evening up, the final conflict is looming, and then slaughter all your characters, crushing their goals. If you are writing a tragedy, you need to let your readers know it's coming. By the midpoint of the story you should have several clues and a good bit of foreshadowing in place to let them know things are not going to end well for Team Good. In fact, though I'm not generally a huge fan of the preface/first chapter that shows how the story ended, if you've ever been tempted to use it, when you're writing a tragedy is the time.
Otherwise, set up the plot, even up the power balance, and to quote a wise man, "Have fun stormin' the castle!"