Have you ever seen an old map with large stretches that say: “Here, there be dragons”? This was often done because the cartographer didn’t know what was there. When you are creating your own world, it’s a sign that you haven’t decided yet.
In this writer’s opinion, one of the most vital skills for a gamesmaster (GM) is improvisation. No matter how well you’ve prepared, there will always come a time when the players ask a question that you don’t have an answer to, want to take an action that the rules (as written) don’t allow for, or wander off your carefully-prepared map to look for the dragons. At that point, the good GM improvises an answer, rather than shutting the player down.
This idea might be daunting at first, but the flexibility that a GM can provide (rather than a static computer-based roleplaying experience) is a good thing! It allows your game to be truly interactive and for the players to take whatever action they like, rather than choosing from a list.
Different people start with different levels of skill and confidence. Don’t worry — the latter is more important than the former, and both will develop with time and practice. If you don’t have the confidence, though, the skill will never get the chance. So, how can we become more confident with improvisation?
For me, there have always been three main strategies. The first is having the trust of your players — a topic which could be a whole article by itself. Most players will be looking for consistency and fairness from you. In many games (though not all) the GM has far more power over the world and its inhabitants than a player does. Every action needs to be agreed by you, whether implicitly or otherwise, and you are making decisions for the myriad non-player characters (NPCs) that populate the world. You control the ultimate decisions on how the king treats the party, or how fair-minded the Galactic Order might be — and the players can only respond to that.
There is an old GM joke: “Rocks fall”. It refers to the power that the GM has over the world and the idea that if a player annoys them, they can declare that rocks fall on that player’s character (or the whole party) and that death occurs. The story may be apocryphal, but it still makes an important point about a very real power imbalance. Trust comes from not abusing it.
Does that mean that every combat or encounter needs to be perfectly levelled so that the players can win? No, not at all. Some groups want an adversarial GM or prefer a setting where not everything is level-appropriate, and they are liable to run into threats that they can’t handle. What it means is that those threats should be foreshadowed, and players should get a reasonable idea of what they can expect. If the king is a tyrant liable to throw them in a dungeon at the first sign of disrespect, there should at least be an opportunity to learn that about him, if not actually witness examples. If a huge dragon lives in the local swamp, nearby residents should know about it.
You also need to ensure fairness between players. There is a reason most games have initiative systems: so that each player gets their turn, but you need to ensure that each player gets their turn in other scenes too. Outside of combat I try to ask each player what they intend to do before resolving any actions, to ensure that they all have an opportunity to act. Your calls on the rules also need to be fair; if you would let one player do something, you need to extend that to all other players, and this can be tricky. Unconscious bias is a very real threat.
All of this should help your players to trust your decision making and, as they trust you more, so too will you begin to trust yourself, which will help your confidence when called upon to make those choices.
The second strategy is good knowledge and understanding of your subject matter. Ideally, this will be greater than that of the other people playing, but that isn’t strictly necessary, or even always possible. You don’t necessarily need to know your setting and rules verbatim, but you do need enough awareness to give you an idea of what is reasonable in different circumstances. At the least, try to know what the options your players have invested in do, so that you don’t undermine those choices. For example, a player who has chosen for their character to be a wizard specialising in fire spells is going to be disappointed if anybody can better their fireball by throwing an oil flask.
The same need for consistency holds true of the world and setting details. If there is a war going on between two nations and the party works for one of those, they are unlikely to find help from residents of the other. Note, I say ‘unlikely’ rather than ‘impossible’. Keep your setting in mind but don’t let it dictate your decisions if your players do something unexpected, such as seeking help from ‘the enemy’. Perhaps the person they are approaching is a double agent, or a veteran who has seen enough of war and now tries to help anybody in need. There are lots of justifications that can be used, but you need to know your setting in order to prepare them.
The better your knowledge, the more likely your decisions are to make sense, which itself supports the consistency and verisimilitude of your game and bolsters your players’ trust. You are looking to make decisions that are in-keeping with the spirit of the original rules, setting, or adventure. The ideal situation is when your players can’t tell that you’ve started to improvise rather than work from your existing knowledge; the more seamless this is, the more believable and authentic a game you have.
The third strategy is, quite simply, to have some tricks prepared. Sometimes you might not be able to think of an answer right away if the players have done something that really surprises you. However, there are ways you can play for time. You’ll probably want a drink with you during the game anyway – a lot of talking will dry your throat out. Having a small glass can be better than a big one because, if you need to play for time, you can finish your drink and go to get another. A toilet break can provide the same opportunity, or simply say you need to look something up and go for your book.
Why don’t we go straight for the book? It will, generally, take too long and, in many circumstances, a quick decision is more important than a correct one. If you make a mistake, you can fix it later, or discuss with the party if you think your mistake might actually be better. Either way, you’ve kept things moving, which is the goal — delays break immersion and pull your players out of the game.
There are other useful tools that you can keep to hand. Naming characters can be the hardest part of creating them, especially when, as GM, you need to create so many of them; so a list of names is a fantastic resource (internet name generators are not the answer, unless all your inns are cheap nasty holes that rent by the hour, and none of your NPCs are ever to be taken seriously again). Alternatively, you can have a system — such as naming alphabetically. If the first NPC had a name beginning with A, the next begins with B, and so on. Limiting the options available to you can help you come up with something. Maybe all characters from a certain place in your world have names that sound like they come from a particular region of ours (though try to avoid the often slightly offensive typical stereotypes that might go with those regions).
Ultimately, though, the most important piece of advice I can give is to say “yes”. Unless there is a burning reason not to, always try to say “yes”. Is there a chandelier? “Yes, there is”. Do I know anybody who works at this space-station? “Yes, you do”. You might put in a “but” — “yes, but he doesn’t like you very much” — to make more game from it, but a “no” shuts things down. A “yes” lets players proceed, and the player with the idea gets to be awesome. A “yes, but” progresses, and with a complication that the players now need to resolve — subplot!
All of this is well and good for the ongoing decisions you need to make throughout a session, but there is a nightmare that we haven’t discussed yet. What happens when you have no idea what to do for the session at all? It isn’t answering a single question that you need to worry about; the doorbell has rung, the players are here with their dice, but you’ve got nothing. Can you improvise your entire session?
Well, you could try to avoid the issue entirely (some ideas for how to do that were presented by Johnny Nexus in Critical Miss back in 1998 — though beware, some of the humour may not have aged well). Alternatively, you can improvise something, and it’s amazing how far you can get with not very much. Remember that you don’t necessarily need every answer until somebody asks the question.
You do need an opening concept, which is probably the biggest challenge. Easiest to fall back is on the McGuffin; a plot object (or person) that everybody wants for different reasons, but not everybody can have. Either the party’s superiors (or clients?) want it, or they want to stop somebody else from getting it. Doesn’t matter what it is for now; you can keep it vague if the information is rumour (Lord Evil has lots of men around the old tomb — he must be after something) or prophetic (our greatest sages say that our salvation lies at the centre of the Forbidden Forest — you must find it and work out what it is on your way).
The setting of the adventure will probably suggest encounters to you; forests will feature wild beasts, faeries and similar. An ancient crypt is probably full of undead and curses. A wizard’s tower will probably have golems and the odd demon. You get the idea. Many games will provide some guidance on monster stats and how challenging they are, so you can pull together something suitable reasonably quickly and easily.
Try to make sure that it isn’t all combat; the players of the less combat-focused characters need something to do. If you can come up with a puzzle, that’s great. If not — don’t worry. Present a problem, and when the players come up with a reasonable solution, let it work. If the game allows for it, make it into a skill challenge or similar. Perhaps a chase could liven up the end, especially if the characters manage to get whatever it is first.
And as for whatever it is — don’t sweat it. The players will discuss that between them in the gaps between encounters (whilst you are pulling stats together). Keep one ear out. I can guarantee that your players will come up with something far more interesting between them than you could, so let them. And then the player that guessed that one gets to be smug about it, too!
Improvisation, whilst important to learn, can be a very scary part of GMing, especially when first starting out. However, it doesn’t need to be. And when you get used to it, you might start to lean on it almost entirely — especially if your players often stray off the map and search for dragons.