On Saturday I had my first go at running
A Wilderness of Mirrors. It's a poster-child for the Indie-game movement in a lot of ways: it's self-published, relies on shared narration, is explicit about how gameplay should be, and includes non-traditional game elements. It sounds ideal for a convention game too, because it can be run either in a highly compressed timeframe (
eyes_of_winter ran it in an hour later in the con) or be expanded to fill a longer time, and because it's pick-up-and-play.
The game works in 3 phases:
1. Create characters
2. The GM pitches a one-line mission, and the players detail that mission
3. Execute the mission
Mechanically characters are very simple: they are rated in the 5 classic spy activities. Leading, Talking, Gadgets, Sneaking and Killing. These are bought on an inverted triangle of cost, encouraging people to specialize heavily.
I understand why the characters are so simple, but what happened for us, was that we had pairs of characters that were very similar: distinguished entirely by how they were played. This shouldn't be too great a problem, because that's all that distinguishes most characters fundamentally. The game is built around the spy-team structure, which does help somewhat, because even though two characters might have the same maxed skill, they conceived of themselves as different parts in the spy machine.
Generating the nominal story is equally easy. I pitched a mission, and they make up how the mission will be accomplished. For each thing they need to do, I award bonus dice to be used in completing the mission.
This is where I first started to think things were not going exactly as the game writer intended. Like most roleplaying groups that need to make a plan, my group were good on the big picture stuff and the flavoursome descriptions. It took quite a few pointed questions to focus them into a more detailed plan: one with risks and rewards. Slightly problematic too was that rather then trying to figure out how to execute the mission I'd actually given them, they kept trying to push out the mission to see what would happen next.
The mission was "Execute this English Cardinal, and recover from the body a stolen cypher". By the time the plan was finished, they had to do this on a time-frame controlled by him meeting with Spanish agents, and not only was he to be killed before then, but they had to have someone take his place at that meeting, then kill the Spanish and steal their ship to travel back around the coast and up the Thames.
After the plan, we didn't know exactly how the Cardinal was guarded, what his habits would be, or a host of other useful details - almost 2/3 of the items on their plan occurred after the mission I'd given them would be over. This is no great problem: it's not like I'd spent a lot of time writing a pre-gen adventure I'd now have to toss.
But there were a couple of other things I didn't pick up on quite as quickly. The plan wasn't especially risky, or detailed. By which I mean that while there were a sequence of things they planned, it wasn't clear to me what the timeframes were, what could prevent the planned actions from taking place, and what the possible consequences might be of a failure at each stage. Perhaps a slight pause after the plan was finished would have enabled me to think of some things, but 2 hours for the complete game does not allow much thinking time.
And then it was time for them to play through. Right from the get-go I tried to create trouble by doing things unforeseen in the plan. Having their contact in Edinburgh under surveillance, having the Cardinal move up the time-frame by scheduling a public appearance, and that kind of thing. I relied in several places on the players to narrate the details of what had gone wrong, my brain not being quick enough with the required details.
Due to time constraints, we had to finish after they'd killed the Cardinal, but before they'd moved on to their post-assassination part of the plan.
Some general comments on issues I noticed.
The basic mechanic is (Skill+Expendable Pool)D6, but most characters did their damnedest to operate within the scope of their 5-rated skill. To get total narrative control of a scene outcome, you need to get a total on your dice of 20, so 5*3.5=17.5 implies that you need only a marginally above-average roll to get total control of the scene. Allocating a single Mission Point die solves that problem, and the odds of rolling a total below 5 on 6 dice (5 is the cut-off for GM narration) are, well, pretty low.
This in turn meant that once the players were on their story track, it was difficult for the GM to have much influence, and so in order to raise challenges you need to obviate all their planning (i.e. effectively ignore the first half of the game)
There are suggestions for arbitrarily reducing their competence as time wears on to put them under pressure, which I didn't properly implement. But if I had, then the table would have turned and in the last few scenes nothing would have worked for them. A better way is needed than presented in the book IMHO.
My group decided to effectively act in concert right up until the end, where the two Malcs fell out. There is a mechanic for the character to betray each other... but there isn't really any motivation for them to do so because the pay-off is additional dice, which as I've noted above, are not terribly useful because it doesn't matter whether you roll 20 or 30 you get the same narrative control.
As a couple of hours of gaming goes, this wasn't bad by any means. I thoroughly enjoyed the game, and it ran just as quickly and smoothly as I'd hoped. But I found myself disappointed that the game mechanics seemed to militate against the sense of tension and uncertainty that pervades my favourite spy fiction. "A Wilderness of Mirrors" suggests a landscape where you can't see where you're going, where you can't be sure that what you're seeing is real, that you need to be vigilant and alert... and I got very little of that from the rules as written.
This is a very promising game, but very far from a finished product. The basic shape of the game is really appealing, but there needs to be more thought put into introducing risk into the plan, and into sowing mistrust and betrayal amongst the characters. I think that hybridizing the Mountain Witch's Dark Fate mechanic in some form would work well, and widening the margins for each success level would allow the GM greater input and monkey-wrench application without having to blatantly derail the PCs plans.
And having written that, it occurs to me that with an appropriate hack of the Mountain Witch you could probably do everything that this game wants to do but doesn't.