At first I just liked the idea: the ref doesn’t roll any dice. I don’t know why, but knowing myself as I do I suspect it was just the novelty of it. I frequently ref games and I am always rolling. What if I didn’t? Was there a reason I had to roll?
My first experiment with player-side rollers was in an early test of some ideas for a second edition of Diaspora. I did it in the most obvious fashion: rather than have the ref roll for non-player characters, the ref would just set a difficulty value. It bombed. It was boring as hell. I shelved the idea.
Then I played some Powered By The Apocalypse games. And it works there. Hmm. What’s the difference?
Well one difference is deeply systemic: in Fate the game works by simulating everything with the same structure. Characters are roughly the same whether they are in the hands of a player or the ref (don’t get me started: yes the ref also plays, but let’s acknowledge that they play differently and that this age old set of terms remains sufficient, however imprecise) and so it makes sense that everyone engages these characters the same way, including the ref. PbtA games don’t do this–the system is asymmetric. Opposition does not have the same systemic model that characters do. So with Fate I was expecting to have some fun as ref in the same manner as the rest of the table because that’s a thing Fate does. When I switched to player-only rolling I lost that expected fun. And there wasn’t much left for me because the system doesn’t provide any (because it’s not designed for this, not because it’s bad).
When I got player-side rolling working I found that the ref experience is fundamentally different. Mostly I realized that because I was not playing any characters (in a dice-rolling context, I mean) there was no pressure to be adversarial: there was no contest for me to try to win, whereas when I am piloting non-player characters on a combat map I am certainly playing to win. I make good tactical choices intentionally. I am trying to kill the player characters but with opposition that I have carefully crafted to be unlikely to succeed. Once I framed that in my head that way I realized I didn’t want to do that any more. It’s not something I want happening in a game, generally, as a design goal and so as a side effect it’s not desirable.
But there’s still the problem of having something to do. Absent the adversarial play in Fate, I was bored. But not in the various PbtA games. Why? I wasn’t sure and, as is my wont, I didn’t think very hard about it. I went straight to my own design. This is a deep character flaw of mine: when I play a game I like I generally start my own project and reconstruct elements of that game without thinking too hard about them. Instead I concentrate on thinking about how my new game works. My Fate games are games that I invented after a read-though and play of Fate — an imperfect read through and play. They are emulations based on my memory of what Fate is supposed to be and not literal Fate games.
And so with this Apocalypse thing.
So in Soft Horizon games I built one dominating thing for the ref to do: manage risk. The creative input of the ref is not adversarial management of threats but is instead the creative interpretation of risk. We use Rob Donohue’s risk list (because it’s brilliant) and it turns out it’s filled with creative opportunity. So, for example, let’s say (and this is a real example) characters are trying to escape from a bad military defense: the enemy is overrunning the defenses and it’s time to get the hell out of Dodge. Players grab a motorcycle with sidecar, pile in their wounded, and head for the hills. Or rather, for the nearest city. We determine this will be a CHASE roll with risk DELAY. If they fail, we will interpret DELAY as getting captured. If they succeed but realise the risk, we will interpret DELAY as getting lost. That’s the first part of my creative load and it’s plenty fun but not adversarial: I’m not manipulating their chance of success, just setting the consequences of failure.
They succeed with risk: lost. And so I narrate a scene with them getting lost and we come to another check: trying to find their way. And that moves us in another new direction, again with my creative energy going into developing the risks into a coherent narrative.
One of the richer risks in this system is REVELATION. This will reveal some information that makes things worse. This is your plot twist. But you can’t plan it, because it doesn’t even come up until there’s a roll on the table and it needs to be in the context of the scene we’re rolling for. So as the ref I need to ad lib that revelation — if I had any plans, it’s a spanner in my own works. It’s as much a revelation for me as for the players!
Another example perhaps: our escaping trio reach an old Tomb dig site that seems to be abandoned. They are suffering from sunstroke and one of them is injured. They are ostensibly slave-runners: they hunt down cities that keep slaves and disrupt the institutions that make that possible and free slaves. The nearest city, Morgenstern, is a place they often bring freed slaves to be safe. So in that context, they are making a MISCHIEF roll to break into a supply shed and I set the risk as REVELATION. My creative input: if this risk gets realised then in the shed will be evidence that Morganstern keeps slaves. I didn’t have that planned. In fact it wrecks everything. Awesome.
So that was the barrier and the solution: if as ref I am not participating in the adversarial simulation, what is there for me to do? How do I get to play? And the answer in the case of Soft Horizon is that I need to continuously create, to ad lib. And there is a mechanism that forces me to do that, that constrains what I will need to ad lib about, and that provides cues for how to proceed.
Well it turns out that’s something I like doing a lot.