Quick summary: if you play tabletop RPGs, of whatever flavour, you need a copy of this book. It’s that good.

Like many roleplayers, I’ve often described my hobby to straights as something like improvisational theatre or improvisational storytelling. It sounds a little cooler than pretending to be a knight or wizard who gets to kill monsters and find treasure. Like many roleplayers, I figured mutual make-believe, randomised with some dice, was the core of it.

Enter Graham Walmsley. Graham’s a roleplayer with experience in improvisational theatre, which he draws on extensively in his game-improvement manual Play Unsafe (I’m avoiding describing it as a GMing guide because it isn’t one; game-improvement manual is the closest I can get to describing what Play Unsafe tries to do).

Play Unsafe openly draws on the work of improv guru Keith Johnstone, adapting his stage work to the tabletop. The result is a style of play I think I’ve been fumbling towards for years without getting quite to the heart of the matter. I think I’ve come closest in one-shot adventures, where I haven’t invested huge amounts of time and energy into plotting.

The basic idea of Play Unsafe is to avoid too much planning and go with the flow. Listen to what other players are doing, and build on it (and likewise, trust they’ll build on what you’re doing).

Key concepts are to try to be ordinary, and go for obvious responses, trusting that plenty of obvious events will build into a unique story. Avoid blocking other players (by countering their actions or rendering them ineffective), but carry their ideas forwards – or sideways if you prefer.

Play Unsafe’s ideas have obvious advantages for storytelling games (narritivist games, in Forge terminology), but I think have useful application in simulation games as well. I have to add here that at no point does Graham use Forge terminology; GNS Theory and the Big Model are never mentioned.

The book is divided into five chapters: Play, Build, Status, Tell Stories and Work Together.

In Play, Graham points out how much work is involved in gaming, particularly for GMs. His advice, amply illustrated with examples, is to stop planning so much. Hold ideas lightly is the mantra: have a rough idea of what kind of thing you want to do in the game, but don’t force it (it’s a little like the ‘broad plans’ I discuss in my GMing essay The Illusionist, but Graham carries that idea further than I did). Here also is his advice to be ordinary, boring and obvious – the idea here being that when we try to be ordinary, we will sometimes be inadvertently awesome, yet when we try to be awesome we’ll almost always fail. Being obvious is one foundation stone of spontaneity – and what seems obvious to you may be very cool to others. The other main foundation is to drop your guard: do your best to be honest and avoid censoring your play – when you censor, or let doubt creep in, you lose the moment.

In Build, Graham discusses how to build on other players’ ideas, different ways to work with them, what happens when we block other players’ ideas, and then discusses times when we should, perhaps, use a block (the important thing here, of course, is that we are doing so consciously and have good reason to do so).

Status is a fascinating chapter, built solidly on Keith Johnstone’s theory that all drama concerns status, or more particularly, change in status: the prince becomes a pauper, the servant girl a queen. Graham discusses how to play both high and low status characters with a nice little acting lesson, then discusses the fun you can have playing either. He also mentions a couple of theatrical truths: rapid changes in status are comedy; long-running, slow changes in status are drama (for example, King Lear’s slow descent into madness).

Tell Stories discusses broad story-building techniques – and, as Graham points out, even the most stereotypical dungeon crawl has a story: you meet in a tavern, travel to a dungeon, avoid traps, kill monsters, meet a powerful monster, kill it, and off home with treasure. But stories can be made more interesting with a few solid techniques: build a routine, then upset it. Create a status relationship between two characters, then upset it. Build in mystery. Deliver on promises (one of Graham’s examples: if you mention a prison no one has ever broken out of, you absolutely have to stage a breakout from it later). Present players with moral dilemmas – for more interesting than simply solving a puzzle. Avoid jamming in a combat scene when you’re out of inspiration (there’s one of my great failings as a GM). Get straight to the action: if you’re going to kill the king, do the thing; don’t go on endless sub quests first. Then work out when happens after you kill the king – other players will help you work it out (I suspect this resonates with a lot of Harniacs who play in Kaldor – we all talk about what’ll happen with the king dies, but most of us are too scared to do it).

Work Together is about trusting other players and having fun and being polite and throwing them curve balls every once in a while. It’s about pushing your boundaries, letting go and having fun – and doing all of it nicely, because you’re part of a group playing a game. We all have fun.

What’s notably absent in Play Unsafe is anything resembling a traditional GMing guide: types of players and how to deal with them, setting mood, getting inspiration for plots, how to make your planning more efficient. It just isn’t that kind of book. It’s about changing your attitude to games, rediscovering the fun. And it’s for players as well as the GM, because to make Graham’s style of play work, everybody has to be in on it.

All in all, I think Play Unsafe is the best book of gaming advice I’ve ever bought. It’s cut to the core of what’s been bugging me about my own gaming style for some time, and it’s offered me a different way of doing things. I can’t wait to try some of these ideas in play outside a one-off.

Play Unsafe is available as a PDF ($8) or printed book ($15 + $5S&H) through the Indie RPGs Un-Store.