Twists are those things people tend to love or hate. However, when done well they can do amazing things for a story. As a result, I hope to go over some of the important things to think about when plotting twists in tabletop role-playing game campaigns. Hopefully someone out there will find it useful.
Do Your Players Like Twists?
If your players really don't like twists in movies, books and other things, I'd be weary or at the very least careful with plot twists. It's that idea of trying to give players an experience that they enjoy. It gets a bit trickier if the group is split on plot twists. In these cases, I find a few carefully executed ones tend to go over well. It also helps to know your players. I've seen people who hate the “Outer Limits” style twists at the end but are fine with twists in the middle of the story that make sense in retrospect (back stabbed by that one guy who, looking back on it, makes sense).
Too Many Twists
I find it's easy to get too twist happy. If your players like the hilarious insanity of twists within twists, that's fine. However, if it unintentionally sneaks up on you, that's a problem. I'd focus on a few (around 3 tops) main twists that might occur through the campaign. There might be situations where a spontaneous twist will arise and it is here where things will need to be reeled in. The power of a twist, in my experience, partially comes from expectations being broken. If the players expect a twist, it loses some of its power. For that reason, not throwing a twist every opportunity you have is for your own benefit.
Scale of Twist
I briefly touched on this, but the size of the twist at play could play a role. All right, so your good friend turns out to be a good vampire. That might be fine (assuming it's not an evil vampire pretending to be good). The rest of the campaign is still fine since it doesn't paint all of the events up until now in a new light. It also probably doesn't change the course of the campaign and force the party to re-evaluate their decisions (though it could, and if it did it would be a bigger twist). However, if it turns out the players were working for the bad guy (quite common plot twist), their achievements before this point start looking like losses for the good guys. Some people like this, some don't. At the very least, thinking about how much is affected by a twist is important. I've seen situations before where a twist had quite large rippling effects through a campaign. It lead to some great fun, but not all Dungeon Masters like that level of unpredictability brought on by themselves (they prefer that the unpredictability comes from the players).