One of the most important skills in the Dungeon Master’s toolbox is the ability to improvise.

And, based on my long experience with the game, it’s a skill that takes time to develop. Like most skills, you get better at improvisation by actually improvising, and paying attention to what your players think of the result. Many of the games I’ve run have been improvised. Even when you prepare material, there’s likely to be some improvisation involved. After all, it’s the key thing that distinguishes role-playing games from other games like Chess, where the rules are cut and dried. However, the amount of improvisation you use in a session will vary.

The advantage of preparing for a session is that you can get a better grasp of what is available for the players to do. You can think about various avenues of approach beforehand, and try to anticipate what choices the players might make.

The advantage of improvising a session is that the players aren’t bound by your preconceived ideas of where they will go. The flip side of this is that you might need to make up a lot of stuff. And continually coming up with good stuff – or any stuff at all – is hard work. Honestly, for many DMs, you’re better off preparing rather than just relying on improvisation. You’ll need to improvise anyway, because that’s how the game works.

And beginning DMs are unlikely to be that good at it. That’s based on my experience. Thirty years ago when I began DMing? I was terrible!

You get better at improvising and design the more experience you have. What kind of experience? Any kind! Of course, doing a lot of improvising helps you get better at improvising, but simple life experience gives you a huge resource of things that you can incorporate in your game. Every book you read, every person you meet, every job you do – it can all be used when improvising an adventure.

There are two pieces of advice I can give you:

#1: Trust Yourself. Pick an idea and go with it. See where it leads. Elaborate on it. Don’t second-guess yourself.

#2: Pay Attention to your Players. If something isn’t working, the reactions of your players will tell you about it. Don’t exclude them from your consideration as you proceed.

Of course, those two directly contradict each other at times. Can you trust yourself when your players hate your idea? Well, yes, you can. And yes, there will be times when whatever you do, things go badly. Can I run terrible sessions, even with thirty years of DMing experience? Absolutely I can!

About a month back, I discovered that only one player had turned up for my regularly scheduled game of Curse of Strahd. Meanwhile, another table had just collapsed – and we hadn’t yet organised a replacement DM (or players). So, instead of just cancelling the game entirely, I decided to improvise a short adventure so everyone could enjoy themselves. After we finished, I decided to write it down – the result got published on the DM’s Guild today as The Witch of Underwillow.

I knew, going into the session, that it was for a group of characters that would likely soon be playing Curse of Strahd, so I set it up in that milieu, with the characters having just arrived in the village of Barovia. It needed to be a quick quest, so I knew that a quest followed a particular structure: a hook, the challenges, the final encounter. The hook? Let’s have a baby kidnapped by wolves.

So: the characters investigate, find the trail, and set off along it. They meet wolves and fight them. Then they get to the place the baby was taken: the witch’s lair. Witches make good low-level opponents in Barovia. I wanted to make it feel strange and unearthly, so I described how the path led underground, under a willow tree. Good unsettling stuff.

I set a couple of challenges – not directly fighting challenges – that I wanted the players to accomplish before they fight the witch. For one of them, I took as inspiration the Knights and Knaves puzzle. That’s where you have two guards, one of whom lies and one of whom tells the truth. Then I told the players that both of the guards are liars. And got amused by the results. (I’m pretty pleased with the solution to that one).

Finally, I was prepared to spring the final twist on the adventure – what the witch was actually doing. I had a plan. I was all ready to reveal it… when they attacked and immediately killed her. Oh well. There goes that idea… I had to make something else up. Such are the perils of improvisation and role-playing games in general. At least I could use the idea in the adventure!

The point of the story is that I improvised the adventure primarily based on my previous experience of adventure structure and encounters. (The two guardsmen came directly from Doctor Who: The Pyramids of Mars, though I’ve encountered the puzzle many times since then). I just put my own spin on the material. Improvising doesn’t mean you have to be original all the time. Get your inspiration from anywhere you can – and then twist it slightly.