Gearing the encounter to something the PC’s want is also important. It always annoys me in video games, and the carry over into roleplaying games, that a lot of the exp gain, or advancement of the groups power level (only relevant to some game systems. Yeah DND, I’m looking at you.) is based off of creatures or monsters attacking the party with little reason. Putting a reason behind it, or intertwining the encounter with other events going on with the party, gives it so much more depth.

Fortunately for us GMs, it’s a rule that players break often — hell, I break it all the time when I play, even when I know better. Mining those “violations” for improv or future adventures is an excellent idea.

Instead, I make notes when they say things. Then I use them for later sessions, preferably with some thought and twisting so that it isn’t obvious where I got the idea. The act of me writing down something when they discuss makes them nervous, and causes imaginations to run wild. Then they find out that the actual situation was somewhat different (whatever I originally had planned), and thus they decide that their comments are “safe”.

I’ve built whole adventures on a single comment delivered during a previous game. For me, these comments are far to valuable to waste by changing the current session.

The best time was when two factions emerged in the players concerning the motivations of a minor NPC. Both sides were wrong, and both of them had very good ideas for what this NPC was doing. Their ideas were far more entertaining than mine. I played the NPC straight. Since this didn’t resolve the issue, both factions stuck to their guns, causing some great roleplaying within the group. I let this percolate for a year or so (real time), *then* I let the fact that the PCs were paying so much attention to this minor NPC lead to other, more powerful NPCs showing interest. The minor NPC got recruited by an order of druids and a resurgent goddess. *Now*, she is important, and in a way that is different from either faction’s view. They were still arguing about her in the campaign’s final session, when it finally got resolved. No way I’m giving that up for the instant gratification of making her more interesting the first time a player’s mind gets tricksy.

A. Some GMs are better served if they “prepare to improvise”. It can be as simple as stock NPCs written up. Heck, sometimes just having a list of names is all you need to keep going. The main thing is that you prepare stuff that interferes with *your* ability to improvise, and that’s somewhat different for everyone.

B. If you are running the kind of game (or your GM style is such) that you want no fudging once the action starts, then you *can* still improvise. It’s only that you have to write down a few notes or be strict with yourself when you keep them in your head. For example, I improvise a fight with a couple of D&D 3E ogres with barbarian levels and exquisite manners. Doesn’t mean that I stop for 15 minutes to do a complete write up or that I make it all up as I go. I take 30 seconds to think about ballpark AC, HP, weapon and armor, and Diplomacy skill, then factor in an appropriate amount of rage. It doesn’t have to be exact, canon D&D–merely in the ballpark. Then I write it down. That’s really no different than if you did the writing down before the session, as the GM is free to make up odd creatures.

As you game, occasionally take note of the players conversations about what’s coming… their conjectures are sometimes far more creative and appropriate than your own ideas.