SHADOW SAINT update – I signed 500 copies at Goldsboro books, I have a final non-ARC printed copy in my hand, and it’s lovely.
Last week, I ran a Black Iron Legacy -themed poll-driven game over on the twitter machine. I didn’t have any real plans – everything was made up more-or-less on the fly, or inspired by stuff in the novels.
The story of the game, by the way, went something like this. Tyrol was a former saint of a foreign god of mirrors, the Duke Behind The Glass. After losing a battle with a rival saint called Yauro the Golden, Tyrol ended up working as a saint-hunter in Guerdon. He investigated the ritual murder of a theatre student, which led him to discover that there was another saint on the run in Guerdon. This saint, a devotee of the god called Smoke Painter, had actually gone rogue and murdered Yauro the Golden. He’d fled to Guerdon, and had tried to wean himself off worship of the Smoke Painter by mixing prayer with theatre.
Tyrol found this saint and grabbed the heart of Yauro the golden, returning through the mirror and regaining the favour of the Duke Behind The Glass.
I’m unsure how clear that was to the players – one of the joys and frustrations of roleplaying games is that the player may never learn exactly what’s going on, and has to make decisions based on incomplete information. In this case, for example, the players never chose to follow up on any of the leads that would have led them to the actual killer of the initial victim (another student, Tarantis Voller, who was falling under the influence of the Smoke Painter through his association with the older saint.)
What’s interesting from a game-narrative point of view is that this medium lets us look at individual decisions and analyse how they fit into the larger story. Take the very first poll I ran.
Here, the players are basically picking the story they’re going to get – I had a vague starting premise in mind for each character, but the premises were all wildly different. The players didn’t know what the story was going to be for Tyrol or Pearl or Uriza, but that initial choice defined the context for the rest of the story.
Another critical branch – there’s a whole other story where Tyrol’s digging up dirt instead of jumping through mirrors. Compare, though, to later decision points like:
In contrast to the ‘set the scope of the story’ decisions, this is basically a question of ordering priorities. Making Tyrol a fallen saint as opposed to a journalist or a mercenary cut off a whole universe of possibilities; there’s little sense here that the player is cutting out other options by their choice here. Indeed, I’d argue that the player would feel betrayed if this sort of ‘ordering’ question abruptly booted them into a wildly different situation. If, say, picking “try to identify victim” brought the player off to the city watch station and then into, say, a fight scene, that would feel like a trapdoor. Choices that appear non-exclusive should generally be non-exclusive.
(I had a similar experience playing the fascinating AI Dungeon 2 over the weekend, which uses a neural net to interpret your text commands. Sometimes, it works wonderfully, but sometimes you’ll give a command like “climb down off roof” and it’ll respond “you climb down off the roof and leave town. A week later, you find yourself in a forest being attacked by orcs…” – which is a disturbing loss of agency. I just wanted to try the door on the ground level, not start an entirely different adventure!)
A third type of choice for the player is a tactical choice.
Here, the player is clearly making an exclusive choice – you pick one conversational gambit and stick to it. However, there isn’t the same scope of definition as previously. You know that whatever you pick, you’re going to be talking to Dr. Ramegos, as opposed to choosing between being a journalist and being a saint-hunter.
So, the player’s choices might be categorised as
Declarative: My character is X, not Y
Ordering: I’m going to order my priorities thusly: Y, X, Z.
Tactical: I’m going to do X as opposed to Y
Now, what fascinates me (and, like, three other people, one of whom is definitely Morgue) is that these decisions feel completely different on the other side of the screen. Declarative decisions are basically creative prompts for the GM, demanding I come up with new material on the fly.
Most ordering and tactical decisions are relatively uninteresting to the GM. I mean, take:
That looks like a critical choice, but really, I knew the player was going to get the key clues (he’s an illusion, there’s a magic play, Tarentis Voller) no matter what happened. Those clues have to be found for the story to continue (Core Clues, in GUMSHOE parlance). Maybe you dispel the illusion and search the room; maybe you follow the illusion to meet Voller, but it all leads to the same place.
Branch points move the story in different direction. Something like this is obviously a branch point:
But so are questions like this – from a player’s point of view, it looks like you’re just picking the order of events, but from a GM’s perspective, anything involving the passage of time is going to be a branch…
It’s interesting consider how the story came together. We went from three entirely different tales (Tyrol/Uriza/Pearl) to one particular flavour of Tyrol (fallen saint). Picking the Duke Behind The Glass as Tyrol’s former deity locked in another whole set of themes and symbols – you couldn’t have the same story with any of the other gods.
(By the way, I have no idea who Moon Eater is, and the Duke Behind The Glass only took on meaning in this twitter thread. The other two are better established in the novels.)
You also get lovely moments of synchronicity, like this one which let me play into the whole treacherous reflections plot line….
It also strikes me that this format produced an inverse cone of possibility to a regular tabletop rpg. When writing a tabletop scenario, the writer only knows with complete certainty what things are like at the _start_ of the adventure (“you all meet in an inn, and you’re all adventurers, and none of you are injured or cursed or anything like that”) and things dissolve further into uncertainty and if-then statements as the adventure progresses (“if the players have the golden key, but the gnome is still alive and hostile to the group, then the gnome attempts to steal the key at this point. However, if the gnome is friendly, and the group don’t have the key, then…”) Whereas here, the moment of maximum uncertainty was at the _start_of the process, and I was able to offer choices that progressively closed down possibilities – it’s not that “Tyrol steals the heart and becomes a saint again” was the only outcome, but I had to push the story towards one of the few thematically resonant and satisfying endings that remained in the possibility-space…
After all that, I’m not sure if I should be thanking the players for playing or asking them to fill out waivers for being part of an experiment in ludic narrative…
Edit: Ah, I think I’ve identified the insight I was fumbling towards earlier! Declarative decisions are the most…, uh activating for a game master, requiring the most thought and creative response, but they’re also the ones that are usually taken away by prewritten adventures. The key, then, is either finding places to allow declarative decisions while still providing GM support, or else ensuring that the ordering/tactical decisions available to the players are made especially interesting to the GM…
Gareth Hanrahan 
This is really cool, I wish I’d been aware of it when it was happening! Love the idea of using twitter for these global-scale D&D style games …
LikeLike