Adam R. As a new GM, I sometimes find it hard to apply your Lazy GM philosophy. This is partly due to my own enjoyment and appreciation of detailed prep, but I have come to recognise a lot assumptions in the Lazy GM approach that you don't seem to speak about much because you have developed your approach from the perspective of an experienced GM. As an inexperienced GM, I have found that this style of prep relies very heavily on improvising succressfully at the table and I find myself struggling to have the confidence and knowledge to do this. For example, just listing out the type of monster for a session relies on you being able to create a balanced encounter on the fly. Additionally, you seem to have no need to prepare dialogue ahead of time, whereas I find myself struggling for words and contradicting myself when I have NPCs providing exposition. Can you discuss what the Lazy GM approach assumes about what a GM needs to do at the table?

Being able to improvise is indeed one of the areas GMs can continually improve on and it's definitely challenging. One thing to know about the steps from Return is that they didn't just come from my own experiences but the experiences of thousands of surveyed GMs and hundreds of interviews with GMs. I studied tons of gamemaster books to come to those steps and the advice the book offers.

One of the core principles many GMs come to – including many of the designers of the games we're playing – is that improvisation is a critical GM skill and we just can't know where our game is going to go.

When I interviewed Matt Mercer many moons ago, he said that he was improvising about half the NPCs he ran in his game – and this was at the most viewed RPG games ever done.

One of the things that Return tries to do is push you towards "preparing to improvise" by separating out components of prep rather than mashing them together into "encounters" or scenes – which is typically what GM books recommend. The problem is that scenes just don't play out like we think they will – and that's fine! It's good! It means we get to watch things evolve ourselves.

The two examples you bring up are good points: NPC dialog and monsters for combat encounters. In my opinion, it's valuable to change how we prep these specific components to lead us towards improving our improvisational skills. It doesn't always work perfectly, of course, but that's ok. We're just running games for our friends. Not everything needs to be perfect.

Tools like the Lazy Encounter Benchmark help us ensure we don't overheat a combat encounter with too many big monsters when that wasn't our intention even when putting together encounters we didn't count on.

Another interesting thing is that a lot of popular old-school games like Shadowdark reinforce improvisation as well through random encounters, hostility, distance, and other criteria we won't know ahead of time. So this isn't just an idea for new games like 5e but also how things were done and are done with more old-school style games as well.

Lastly, speaking about dialog for NPCs, I always reinforce that the steps from return are there to serve you. You can add as much material in them as you feel you need like conversation points or key things NPCs need to say. I'd argue that it's better to mix secrets and clues and NPCs together so you don't have to bury a piece of info in the head of one particular NPC. That's how secrets are intended to work.

Above all, the steps serve you. Prep what you need to run your game. But take a look backwards at the games you've run and see what helped you and what did not. Refine your prep to the parts that best help you run your game and omit the parts that don't. We're all different and different games or sessions require different amounts of prep. The steps are guidelines, not hard rules. Use what makes sense.