Play to find out what happens.
It's not, for instance, your agenda to make the players lose, or to deny them what they want, or to punish them, or to control them, or to get them through your pre-planned storyline (DO NOT pre-plan a storyline, and I'm not [messing] around). It’s not your job to put their characters in double-binds or dead ends, or to yank the rug out from under their feet. Go chasing after any of those, you’ll wind up with a boring game that makes Apocalypse World seem contrived, and you'll be pre-deciding what happens by yourself, not playing to find out.
Play to find out: there’s a certain discipline you need in order to MC Apocalypse World. You have to commit yourself to the game's fiction's own internal logic and causality, driven by the players' characters. You have to open yourself to caring what happens, but when it comes time to say what happens, you have to set what you hope for aside.
The reward for MCing, for this kind of GMing, comes with the discipline. When you find something you genuinely care about -- a question about what will happen that you genuinely want to find out -- letting the game's fiction decide it is uniquely satisfying.
Apocalypse World 2nd ed, "The Master of Ceremonies", p.80
I don't know if that's exactly what you want. It sounds very close to what you want. And Apocalypse World isn't the only game you can play like Apocalypse World, it's just the best one. (And you can grab the first edition free here. It's not irredeemably worse than the not-free second edition, just made in a world where Fury Road didn't exist yet.) The second-best one is whichever of its successors you favor the setting for, whether they use the engine (Fellowship, Masks, Rhapsody of Blood) or just the philosophy (Blades in the Dark, Scum and Villainy).
Not all games will be easily or even possibly compatible with the most fun bits, so I'll spell out what those are and what elements in a game engine could make them derail.
The Incomplete World: Clear a Space to Barf
Barf forth apocalyptica. -- Apocalypse World
Draw maps, leave blanks. -- Dungeon World
How do you run a game when you don't know what's going to happen? Well, one important part of it is that you just scatter around a whole bunch of things that could happen. Ordinarily, you might have been hesitant to do that because what if the PCs just spend three hours poking at some purely atmospheric lava pool? But when you don't know what's going to happen, you leave yourself room to listen to what the PCs are interested in and steer toward that.
And of course, in order to do that you need to scatter a bunch of things the PCs might be interested in around the game landscape with wild abandon. Don't commit to anything, don't decide just when you've placed it that it's going to be huge. Maybe the PCs will turn their backs on it, and if it's still small, hey, no loss.
So to start with, focus on sensations. Mention colors, tastes, scents, sounds. Explain things the characters know, but the players don't, at an extremely surface level. I'll give you an example: as a result of a botched roll during setup, my players came back to their rented quarters to find out the place had been drastically rearranged:
And when you came back, she'd reinstated The Dukes. Bodies hung by intricate wire-and-gear-work from the brasswork spheres, now mounted on the ceiling. She's got what appear to be the proper seals from the Last Gatesmen attesting these deaths are by neither her hand nor her malice, same as the first time. Same as the first time, they're slowly being stripped of flesh by her spawnlings, even as they seem to consult with each other and order the spawnlings around. Same as the first time, she doesn't explain why she's doing this - part of her lifecycle, power play, nervous reflex - and you've got no idea quite what it accomplishes either.
This is a game set in something like Planescape, and I didn't want them to be rooming with a bloody-handed murderer, at least not yet, so I came up with something like the Dustmen - the Last Gatesmen, a guild that involves itself in the things that come after death. Through a combination of player curiosity and my own needing to improvise when players made bad rolls (oh, a partial to try and get your living quarters back? She says sure, as long as you can dispose of the bones. Oh, a botch trying to find where to go to do that? Sure, you know exactly where, this needs to go back to the HQ a ludicrous distance away) my players wound up visiting their home turf, taking the occasional request from them, and one became personally involved with them in an effort to hide their necromantic battle suit from its original creators.
But all I started with was the seed. Plant a bunch of seeds and you can see what needs to grow later.
This doesn't work well for: systems where players have strong capabilities of divination. If players can force you to act as though any given part of your game world is strongly and absolutely defined, you can find yourself scrambling to come up with lots of information and attaching it to something that the PCs might not even be interested in.
Prepare to React: The Threatdown
But you are still the GM, right? How do the standard GM responsibilities of session prep hold up when you're leaving a blank spot in the place labeled "end of session"? Pretty well, as it turns out.
During play, you leapt forward with named and motivated NPCs, you barfed forth landscapes and details of society. Now, between sessions, it's time to go back through your notes and create those people, places, and conditions as threats.
Creating them as threats means making decisions about their backstory and motivations. Real decisions, binding ones, that call for creativity, attention and care. You do it outside of play, between sessions, so that you have the time and space to think.
The purpose of your prep is to give you interesting things to say.
-- Apocalypse World 2nd ed, "Threats", p. 106
All of your dangerous people and places and things get their own section in the various offshoots, but whether they're called threats or arcs or fronts they represent the same thing: an opportunity to prepare productively.
This isn't the prep that you might be used to, creating a pre-keyed dungeon and inhabitants that are frozen in time one minute from the midnight of their evil scheme until the PCs show up and the clock starts ticking. Depending on your hack of choice you may not even be prepping notable antagonists, but for games like Monsterhearts and Urban Shadows devising NPCs which serve more to set your PCs on interesting collision courses with each other. If you do prep conventional antagonists, you can set them up to advance their nefarious schemes to present the PCs with motivations to act against them and to give them something to do when the PCs show up.
What you'll be preparing is broader than that: what do they want, and how do they get it? Most AW-derived games will include some suggested activities depending on what role in the story they fulfill - and "they" here can encompass the normal sort of people, but also animals, landscapes, disasters, buildings, vehicles; if it would show up in the opening credits of the TV show, you can prep for it. So what you can present the PCs with is one part or another of this seething mass of desires and actions, and as they assert their own motivations against it, you'll know how it's all going to wind up reacting.
This doesn't work well for: Games with intricate rules for creating antagonist characters. Most AW-derived games don't get any more complicated than hit points, armor, damage, and the descriptions of unique things an antagonist can do. If you find yourself thinking "oh, of course, they'll run into the Motti Gang trying to smuggle shine dust down at the docks" it's easier to just let that happen if you don't have to break for an hour and make members of the Motti Gang with an approximate lifetime of five minutes. It also doesn't work well in games with strict rules for extended recovery periods. Math is cruel, and your PCs may find themselves helplessly watching doom descend because of some very bad luck.
Give Truth, Get Truth: Asking Charged Questions
Once you have the player's answer, build on it. I mean three things by that: (1) barf apocalyptica upon it, by adding details and imagery of your own; (2) refer to it later in play, bringing it back into currency; and (3) use it to inform your own developing apocalyptic aesthetic, incorporating it -- and more importantly, its implications -- into your own vision.
Apocalypse World 2nd ed, "The Master of Ceremonies", p.85
So, you've got your pick-and-mix of setting details lying out there, you've got your prep notes for the various dramatically important elements in the world and the things they like to do, and you go to the players to tie them all together, but how?
Well, you present them with these setting details and elements of the world and see how they react to them, and you do that by asking charged questions.
The difference between a charged question and a more open-ended exploratory one is discussed at greater length in this blog about crossing the line but, in a nutshell, it's the difference between
You finally kick the slaver out of the driver's seat and he drops screaming down the canyon. You brake his buggy to a halt, and only then notice the box strapped into the passenger seat. What's in the box?
and
You finally kick the slaver out of the driver's seat and he drops screaming down the canyon. You brake his buggy to a halt, and only then notice the box strapped into the passenger seat. So, pertinent question: what do the slavers use for barter?
which can be restated more boringly as
You finally kick the slaver out of the driver's seat and he drops screaming down the canyon. You brake his buggy to a halt, and only then notice the box strapped into the passenger seat. The box is used to transport whatever the slavers use for barter. What's in the box?
I've pulled the charged (some would say "loaded") fact about the world out, it's in italics. It serves a couple of purposes - it gives the person answering the question something to think about as a starting point for their answer, and it lets them know how I'm going to use their answer. This is a much easier prompt to respond to than "say something cool about the world, I'll tell you what for later".
(It's calendar scrip, by the way. Their concrete compound used to be a printing concern, and they've stamped pages of that old product as promissory notes for a day or a month or a year's unskilled hard labor.)
Now, you might get some pushback there, and it's important to listen to players who might object to the facts you introduce about the world, but ultimately you're still the GM, still curating their input, still responsible for running the world in their absence. If you want water to be a concern in the post-apocalypse*, then when you say "how low does the water tank have to get before you start getting concerned", assuming both a water shortage and a water tank, you might get pushback. Someone might say "what? don't we get it from the river?" but one of the cores of post-apoc fiction is you ain't got enough, so you can incorporate the idea that there's a river and say: sure, on days when it's running clear and doesn't smell too bad it's kind of a rush to risk cupping some up in your hands, but otherwise you've got to boil it and filter it and it goes in the tank. So how low does the tank have to get before you start getting concerned?
* Yes, I'm aware of The Waterbearer. Obviously you don't contravene the playbook that exists entirely to explore the interesting circumstances surrounding the guy who always has water in a thirsty world. But let's assume nobody wanted to be one.
This doesn't work well for: Mostly this is about player types, but there are certain systems that attempt to model the world in such detail that something you intended to be an open-ended question with an answer that only mattered for the story actually has important mechanical implications when you get curious and go look up what it means. One type of player it doesn't work well for is someone who's used to those kinds of systems and will be worried that even though you seem to be asking a simple question, they might be screwing themselves over by giving "a wrong answer", and will try to dig into implications you didn't intend. Another is someone who shows up to game night more to socialize than to play, to appreciate the story that's going on around them without contributing much. Some systems can incorporate a player like that into more structured play procedures, but they don't do well when put on the spot to answer open-ended questions.
Run to the end of the world.
So there's your basic plan for playing a game even you, the GM, can explore:
- constantly introduce minor setting flavor that doesn't have to mean anything, so you can catch onto what the players are interested in and make it mean more
- if you introduce opposition, plan its reactions more than you plan its actions so it can more easily fill a space that opens up for it
- use relevant information in the world to prompt players to make connections, then run with them
It's possible that your favorite system might not be able to do this easily or well. It's possible it might not be a good fit for the player group you put together. It's possible it might not make it a good fit for you, since while you can productively prep for this style of play it places greater demands on you in the moment to look for and find interconnections to keep the game going.
But if you can make it work, it's pretty great.