This Sunday, while running a game for my group, I revealed that a recurring NPC the PCs had believed to be an ally was actually a rather sinister antagonist. I believed, when creating his stats, that he would be a difficult enemy to defeat; indeed, I intended for the PCs to fail to defeat him and for him to teleport away through an interplanar portal. A force field blocked off the group (and the NPC) from the NPCs two lackeys that were generating the portal. However, the Warlock in the group used the “Iron Spike of Dis” power. This power, while more than bloodying the NPC, had the effect of immobilizing him. Improvising, I had the NPC open a portal in the floor below him and appear behind the force field and thus out of harm’s way. The Warlock, who was proud of his strategizing, felt as though I cheated him out of his glory. If I hadn’t, I would’ve cheated the entire group out of a potentially epic and rewarding campaign. Did I handle this the right way?
I addressed a similar question a few months back – What do you do when player actions clash with the outcome you expected in the story? I suggest you read that and the discussion that follows, as the point was explored in more detail in the comments.
However, there’s another aspect to this question that I’d like to explore: the challenge to using recurring villains. Recurring characters add a lot of flavor to a story. Looking to comics, characters such as Doctor Doom and the Joker can be as interesting as the heroes they fight. As a player gets to know a villain, his appearance in a story serves a number of purposes. It lets the player have some idea of what to expect (Magneto? We’re going to be fighting evil mutants, guys). Beyond that, the history between the players and the villain can add dramatic tension, especially if the villain has occasionally won in the past. At first you didn’t care about the Lord of Blades. Then he killed your sister. This time – it’s personal!
Of course, in superhero stories defeated villains generally go to jail. As a system, D&D is set up for bloodier battles. In 4E, you can technically decide if an opponent dropped below zero hit points is dead or simply down for the count – but as The Knights Of The Dinner Table have shown us time and again, mercy for enemies isn’t always a top priority for adventurers. You can always come up with a plan for how you want the villain to escape, as with the portal above. But clever players can always come up with a solution the DM hasn’t thought of. So how can you save an NPC without your players feeling like they’ve been cheated of victory?
The first answer is the simplest: You don’t. The underworld has a revolving door in D&D. Raise Dead, Resurrection, Reincarnation… in many stories you can’t be sure the villain is dead until you’ve seen the body, but in D&D you can’t be sure until you’ve burnt the body and trapped the soul.
Of course, with a good story, even a burnt body won’t take someone out. Count Feldvar cut off his finger and gave it to a friend to ensure that the ally would have a piece of his body to raise him if required. The Faceless One is an agent of the Goddess of Death, and Death will never hold her for long. And wily Serella made a bargain with an archdevil—she’s regained her life, but now she has a debt to repay. To me, this last example is the most interesting one because it can actually drive a future adventure. By the time you next meet Serella, your characters will have gained levels. But now she has new powers gained from her infernal pact and a new mission. She’s going to wipe out the city of Ux with a terrible plague because she HAS to in order to pay off her debt – something you set in motion by killing her!
Another long term option is to have a new villain step in. How come it’s only the HERO who gets to say “You killed my father, prepare to die?” Just because the PCs never knew that the villain had a lover/child/parent/partner/apprentice doesn’t stop you from pulling such a figure out for the next story; when Beowulf brings down Grendel, it simply sets the stage for the even tougher battle with Grendel’s mother. Rather than looking at the death as the end of the story, it may be an opportunity for it to evolve in a new and interesting direction.
Perhaps you needed that villain to escape because the story you’re running just won’t work without him. Realistically there’s no time for him to be raised, and you don’t have stats handy for his mom. You just weren’t prepared for things to go this way. You could take the trapdoor getaway approach and hope the players aren’t too annoyed. Or you could take the Doombot option – which is to say, the villain the PCs killed wasn’t the villain at all. As soon as he dies, his body transforms into the corpse of a changeling – but exactly when did the switch occur? Alternately, spells like domination or mind seed could allow the villain to use a dupe as a puppet – and when they get to your final scene, they’ll be surprised when a seeming stranger announces “You fools! Now you face the TRUE Count Feldvar!”
In any case, my gut reaction is: don’t rob the players of their victory. Find a way to make that victory part of the story. Find a way to return the villain from the dead; have a new villain be created by the defeat of the old; or allow the players to have the victory, but have it turn out that it wasn’t actually the villain that they were fighting in the first place.
You know your players, the history of your campaign, the mood at the table. At the end of the day, my advice is to be flexible and to do what is best to ensure that your players have fun.