Counting the fallen just happens when we know a battle is over right? When one side surrenders or flees and the other accepts the surrender, accepts the escape, or falls to butchery or thereabouts.

There isn't a circumstance, seems like, where you'd forgo counting the fallen - where there'd be any benefit at all to skipping that.

(Now I want to set up an escape and pursuit after a fight such that the fallen can't be counted right away. Then what?)

We got together last night, with the intention of running Murderous Ghosts while some of ate and then move on to Durance.  Outcome was the 4 of us played 3 times and nitpicked afterwards.  Josh took good notes on our thoughts, which I hope I won't miss much of.

3 Games: Josh -> Mark, Judson -> Tony, Mark -> Judson.

We were playing with the "pencil in" rules of more violent ghosts.  I'm not sure it helped a whole lot.

I'll note first that we worked out after play the the way that the "involved in the ghost's story" paths are meant to work, once the Player knows how many times they can ask about their interaction, the GM should be on page 10, watching for one of those things happening.  I think every time we actually hit that path, the player busted, and there was a brief awkward scene until we strayed from expectation and got assaulted - the GM didn't have the Interacting with Ghosts choices in mind, and I don't think any of us as players considered them in those scenes.

We definitely hit on "loops" in the graph.  It seemed like there are a few places in the graph of "player draw" node that form a 3-cycle, and where the turn-to directions align with the hands such that reasonable play will lead through the loop.  Does that make sense?  Something like, A) draw: low-hand -> B draw: high-hand -> C draw: bust -> A.  (It's really hard for me not to use graph theory jargon here.)

Part of that was also that the "reasonable" actions of a character were often "run the fuck to the exit," and once the player realizes that that's the only way to not be murdered (or otherwise lose), they tend to align with the "fucking run" tactic.  

Tony sort of nailed the problem we had as a result, and that I recalled reading in other playtest reports: the player's goal is to survive, and is playing against the mechanics.  The GM's goal is to give herself the wigs.  Once we start falling into loops in the mechanics, Tony especially (the rest of us somewhat but less so) got frustrated, and so weren't interested in playing to the wigging.  "I give the giant bloody slavedriver a hug."  That moment was pretty sudden and irrevocable, I think.

Likewise, the decisions the GM makes for the ghosts, and about evidence and how they behave didn't feel significant, since it's the player's job, essentially, to run away from it, regardless of what it is.  

We wanted, as a result, for there to be opportunities for the player to confront the backstory somehow (as well as running away), so that we'd engage more over what the ghosts were like and why.

Final note: Mark seemed to have the smoothest time as GM.  He was flamboyant in his descriptions, and direct in the threats the ghosts represented despite because he wasn't paying much attention to his booklet and frequently was getting off the track with things.  (E.g. GM - 3, described the ghost being in the location still, player draw: high hand, so no ghost.)  Awesome junkie pimp ghost, complete with a separate victim already in thrall.  Result: murdered.