Tuesday, July 26, 2016

Leaving Crowsmantle: A Campaign Postmortem

Last week we had our final session of Crowsmantle. That’s my PbtA hack for our game about adults returning to a fantasy world of their youths. I’d rate myself as generally happy, moderately OK with the campaign as a whole. More than anything over ten sessions it lunged wildly from a rock solid session to a weak one to a great one and then back. My earlier post outlines the ideas we had going in and has a link to the pdf of the rules I wrote. The basic pitch was, In their youth, they journeyed to a land of wonder. There they became heroes and saved lands from a great evil. Then they grew up. Now as adults, they’re called back to fight peril to a realm transformed.” I wanted my system to be light and adaptable.

I posted some sample characters. If you like watching online games, you can see the full series here. I put together a Pinterest board for the campaign. That’s been my go-to for visuals lately. I used to have wikis, but the free ones I used switched over to a pay model, so I’m off of that. Pinterest has the problem of not really allowing for annotation and organization though. I also put together an inspirational YouTubemusic video playlist. I had the players each select a song for their character. I should also say G+ continues to cut event functions. I had cool banners for each session, but because of the way people RSVP now, you end up missing them.

Let me start with the mechanical aspects of Crowsmantle I like hacking games; sometimes they work (Action Cards; White Mountain, Black River) and sometimes they don’t (Scions of Fate). You can see lots of examples of my hacking half-assery on the blog. Many I don’t do anything with, some I get to the table. With Crowsmantle I wanted to try out my idea for a PbtA basis for pickup game aka Pug'buttah. I refined that a little for CM. I had World of Dungeons and Simple World as reference points. Overall I don’t think it’s a bad idea, but I screwed up my execution in several ways: from not having the pdf be cut & paste to leaving aside several of the major systems I wanted to emphasize.

When I showed him the rules, Rich said “this puts a lot of the heavy –lifting on the players.” I acknowledged and yet underestimated that. Rich was pointing to the system for advances. I had offered a general template for them (ala FAE) but didn’t spell them out further. In the 1.1 revision added more examples but it was too late. In effect I gave the players homework, something I’ve said you shouldn’t do. And if you do that don’t expect them to actually complete it. We played bi-weekly, we had a vague set-up, and sessions ran for less than two hours. Instead of coming up with crafted advances, players fell back to taking stat boosts. Nothing else seemed as easy or useful. If I wanted the system to actually work I needed a formula for how you built advances (new abilities, new moves, upgraded moves) combined with easy shopping lists. The examples I had weren’t enough.

That’s especially true with moves. I thought we’d be coming up with custom moves on the fly. But that requires switching modes in play and draws attention to the moves themselves. Maybe if we’d had longer sessions we could have done “move creation” at the end? I’m not sure. I think a shopping list would have been better. And the blame rests with me; I didn’t hold up my end of the bargain. The only move I introduced came in the last session, to abstract and dealing with Mobs of Guards. It’s still an open question whether refined move-making in play would work or not.

There’s a mechanical problem (one among many) with my Crowsmantle rules. I present four basic moves (Fight, Interact, Act, and Discover). Those parallel Fate’s four basic actions. That means you can apply one to any task. If that’s the case, why would you need new moves with rolls? Why should players spend an advance to get a new move if they can just do it with the existing framework? Player-bought moves need to be stronger. That means that the “default” moves need to be a little weaker. For example, my Fight move reads,

10+ Deal take/standard harm and pick three effects.
7-9 Deal take/standard harm and pick one effect.
…deal extra Harm (may be taken multiple times).
…gain +1 Forward for yourself or another (Set Up).
…take no Harm.
…Change State/Position.

A reduced move, let’s call it Conflict, would deal & take standard harm plus two effects on a 10+. On a 7-9 it would deal & take standard harm OR pick one effect.

If a player wanted to buy a Move representing combat proficiency, they could take something closer to the original Fight move. They could also add new choices (Clear a Mass of Minions, Send People Flying Back, Cause Terror in Your Foes, Find a Weakness, etc). That gives a mechanical incentive to figuring out a new move for something your character does well. I’d need to create a better shopping list and build structure to do that.

I had a stat called “Talent” to represent the special magic PCs had when they were younger. It had atrophied. Players named their talent stat, vaguely defining what they could do with it. I planned to use “Big Magic” from Monster of the Week for this. There you roll & get several effects, but you also have a drawback or catch. I tried that a couple of times, but eventually ended up ignoring that catch mechanic. Players got themselves in enough trouble without it. In retrospect I wish I’d had everyone make up a custom move representing their power. That would have helped them define clearly what they could do.

For example, Raven had Futuremancy as a talent, Charlie had Patterns. In practice those ended up way, way too close. Because Futuremancy had an obvious predictive note to it, the group called on Raven to use it, and as a result it got more play. I should have spotted that problem early on. My solution above would have fixed that. Alternately I could have suggested Patterns be explicitly about figuring things in the present scene and Futuremancy be about inquiring about future scenes. Overall I short-changed Sherri in play.

I used Keys as the experience basis, ala Lady Blackbird and TSoY. I still like those, but I’m not certain they worked as well as they could. They fit with our Middle Earth f2f game because they support a certain tone. I’m wondering if something looser and more player defined like flags or directives (from The Sprawl) wouldn’t have been a better choice. Again, it would have asked players to make significant decisions about their characters.

As I mentioned in other posts I’m still getting used to running PbtA. My recent work has split between trad (13th Age, Mutants & Masterminds) and narrativist (Fate, Action Cards). Even when I run the latter I tend to slow things down in combat, extending scenes to give players space for cool stuff. With the latter games I imagine I’m going super-fast. But my ongoing joke with Rich is that after I ran a Fate micro-combat that moved along at a rapid clip, he said “Man, I’d forgotten how long Fate combats take.” Emphasis on long.

I struggled with Crowsmantle combat pacing. I started by following MotW’s approach, which seems closest to my own. There foes have wounds and you have to figure out a Big Bad’s weakness before you can affect them. That went OK, but slower than I wanted. It got better when I abstracted things later on. In the final fight, a hit either took a lieutenant out or put them in mortal peril for another.

My other stumble came from GM Moves in combat. The fight move I mentioned above has your typical pick “take no Harm” under the choices. Early on I got stuck in combat. When players rolled 6+, I fell back to only dealing harm or harming NPCs. Because I wanted speed, I didn’t want to stop off to come up with things. I got over that later. But it’s one of those places I need to work. Fate has aspects and those encourage an interaction with the environment. I think in combat GM reaction moves serve that function. You can use those to cause interesting effects and showcase scenery. I got better as we went along (in places), but it’s something I want to think about more. Maybe I need to write out a set of interesting GM moves for combat scenes. I do like crib sheets.

Session nine’s a good example of a bad session. The previous episode had ended strongly, with players captured as a result of a combat gone wrong. It was one of those great “play to see what happens” results. Since they had been caught and since we had airships in the game, I thought I’d do a Lady Blackbird homage. It landed like a lead balloon; the players either didn’t care about the frame or didn’t catch it. But more importantly dealing with the guards and getting to the flight deck wasn’t that interesting. I had some cool action descriptions, but I ended up bogged down there. The early part of the session sucked the energy out of the room, and we hit a solid meh by the end.

In retrospect I should have done what I ended up doing for the next and final session. I have to give credit to Jason C for making explicit a technique I’d partially used before, but hadn’t pictured in this context. I’m paraphrasing, but he said for a dungeon he just needed a custom environment move, some GM Moves and he could go to town. At Origins (which came after my weak S9) I sketched those out for both Magic, Inc and Neo Shinobi Vendetta. Not just more guards or the opposition is coming. But interesting bits and pieces. I should have had those ready for the ship. Fighting the guards should have happened once, and then shifted to other interesting things.

Session 9 had another problem moment, something which happened several times over the course of the campaign. The players had a choice. They’d made a decision and started on that path. Then that would get obscured, they’d debate the choice again, and it’d become clear that people had missed what was going on. In Session 9, I pushed a choice about direction (go X or Y way with some small structural differences between them) that wasn’t that interesting. I also threw in a secondary event that complicated the choice; I should have just dictated that as my move instead of making it wishy-washy.

Anyway, the lesson I keep learning when I run online is that you have to assume people are only hearing about half of what you’re saying: GM or players. There’s the noise of the medium, user isolation, personal environment distraction, the crunch of a short time frame, and a host of other causes. Having the camera and seeing faces helps, but you’re losing body language signals. When I’m at the table I can tell if Chris is getting it or not. I can tell if he’s even listening. That’s less true online. And if you’ve ever seen players talk over, not listen to, or completely misunderstand one another at the table, that’s compounded online. Some of my Wednesday players are terrible at listening to one another, forcing me to go back and restate points constantly. And that’s a game with a shared map and token space, which usually makes things better since we have a visual representation.

When we originally talked about the game, the group said they wanted to deal with both the Real World and the challenges of The Realm. The idea would be that we’d swap back and forth between the worlds. Once we got into the game, I realized that wasn’t going to happen. For one thing we had four player kingdoms we wanted to at least touch on and ten 1 ½ to 2 hour sessions to do that in. If I’d super abstracted the events and challenges, we might have been able to. But I started more on-the-ground than bird’s-eye. So instead I focused on the other issue we’d talked about: figuring out what had happened during their absence.

I leaned into that because I like running mysteries. That’s one of my favorite things: seeding details, having players come to conclusions, and using those to push things forward. That’s a large part of my f2f games. Assassins of the Golden Age, our riff on Assassin’s Creed, has been pure mystery, problem solving and manipulation. The players spent the last session finding out details and putting them together as a way to take down their enemies. Mysteries work less well online, in part because of that noise factor. But some great players, notably Rich, don’t dig mysteries. He’s not the only one. Judd Karlman shocked me when said he didn’t do mysteries, period, during an episode of This Imaginary Life. Anyway, I spent more energy than I should have seeding details, building frames for mysteries, and leaving threads to be pulled (Eagle Cloaks, Owl Capes, etc). I should have dialed that back and focused on more revelation maybe? I’m not sure. It might be the nature of PbtA, running online, or more likely my own style hang up.

What did I like? That the rules got out of the way. We rarely went back to the book and instead just rolled, assessed, and played. I loved most of the NPCs: the goblin Ortiz, Friend Raccoon, and Subasa all stick with me. Several sessions really clicked for me (the battle at the town and the meditative session afterwards; the capture in the mall; the final episode). Others were seriously uneven. I enjoyed coming up with fairy-tale weirdness (the travelling jazz band; the civilized bear; land shark travel; the Minecraft riff; the Goblin Mall). I liked the world and some of the plots. I thought Mischa’s arc ended up being fairly poignant. I like that Ignacio didn’t really belong in the real world. There’s a ton more. Andrea talks a lot about head cannon. In this case I’ve got a ton of GM head cannon about the setting.


I’ve got a couple of other issues I’ve been thinking about (the difficulty of running a vague vs. unstructured game online, does PbtA work with authorial structures, etc). But let’s leave it at that. If you’ve slogged through my report, thank you. I may get up the courage to do another serious revision of these rules and perhaps even run the same premise with another group. 

Other Postmortems: