As a GM, you’ll often be forced to make snap decisions about an element of the game that might turn out to have lasting consequences — and it’s inevitable that some of those calls will be inconvenient or just plain wrong.

Sometimes you make these decisions because your players surprise you, which is just the nature of the beast (and one of the best things about being a GM). Other times you’ll make a mistake that seems minor at the time, and not discover it until later on.

Depending on your temperament and how important the seemingly-unimportant decision (or mistake) turns out to be, it can be tempting to say “Shit. Now I’m stuck with that” or “Ehhh, I’ll just change it.” When it comes to retconning, though, it’s the specifics of the situation that matter.

For example:

  • One of your players asks how large a building near the party’s headquarters is, and the size you choose on the spot turns out to be problematic.
  • A PC uses magic to learn a secret about a semi-important NPC, and the secret you blurt out doesn’t fit with the story you had planned.
  • Forgetting that the mafia don always sweeps his apartment for bugs, you let the party listen in on a crucial conversation.

In the first case, you can always retcon that decision after the fact. If you start prepping for an adventure involving that building and realize it needs to be much larger or smaller than you said it was originally, just change it — and explain the change to your players at the start of the adventure. No big deal.

In the second case, you’re better off working with the secret that you picked on the spot. There’s probably a way to change something about the NPC or about the story arc you had in mind to incorporate the new twist. I wouldn’t recommend retconning in this situation — chances are good that it will bug at least one of your players. If you absolutely can’t work with what you came up with, retcon it and apologize to your players.

In the third case, your snap decision has changed the course of the game substantially, and the scene itself was probably very satisfying for your players (who came up with a clever idea, executed it and got results). Retconning this one is a terrible idea, and should be avoided at all costs. Just make it work — you might be pleasantly surprised by what emerges from this curveball.

How do you handle situations like this?