The Process of Writing
A couple of years ago, when I was re-emerging onto the KapCon scene after a 10 year holiday, I posted on the old RPGCentral board in Wellington about how people wrote games. My game for that KapCon was a traditional slasher horror set in a shopping mall. The characters were all shift-working night cleaners with undesirable backgrounds being killed off by a pseudo-vigilante.
The concept is pretty basic, and I'm sure you've seen it on your TV screen many times before. My favourite slasher movie is Club Dread, with an affably doomed Bill Paxton. The idea is simple: a bunch of characters are slowly picked off by a serial killer who seems forever one step ahead.
My adventure was just that simple and unimaginative - but on the plus side, when it came to writing it, it was simple and unimaginative. However, I quickly ran into trouble: how do you actually go about putting this paper into a form that's going to be accessible and intelligible on the day, so that if you (god forbid!) have to refer to notes it's not going to be painfully apparent that four or five people are waiting for you to get your shit together? But also more generally - how do you go from having a story idea to having a game that's not going off the rails at the first PC thought?
I sat and struggled with timeline diagrams, flow-charts, and situation flags, eventually chucking all my notes and making a list of horrible ways to get killed in a shopping mall, then working through that list as opportunity arose. Simple, and I guess reasonably effective. All I had to do was obfuscate the timing on any event so that I didn't have any continuity errors. Though if Gino hadn't roleplayed the hell out of his character I'd have walked away feeling like the players were short-changed.
Since then, I've gotten a little better, but mostly at the craft of hiding how poorly prepared I am.
The next technique that I tried to get working, mostly for the couple of years that I was dishing out pulp games, was to write a little story summary. "The players meet X, who tells them blah, which leads them to N." Though basically, I've found that I can plot these gross story movements, I get hugely tripped up in the transitions. That is: once a scene is in play, these little story summaries help me, but I find it hard to get players moving from one spot to another.
The major success-story of this approach was Horror Victorianorum, where I wrote a series of 5 events, then prepared them as hand-outs. I arranged it so that when one character was induced into a hypnotic state, they would get one of these, which would drive the story. I didn't prepare much of anything else, relying instead on my reasonably good working knowledge of Victorian London from running Gaslight, my Victorian Buffy game.
I ran the game 4 times, and 3 times it more-or-less went to plan. The first play-test was a bit rough, because I'd messed up the characters, but the basic plot mechanics unfolded as desired. The only time it went awry was the second play-test, because the players didn't buy into the simple narrative that I'd setup implicitly with the visions: they investigated the hell out of the situation. As a gaming experience it was excellent at my end, as a polished "tunnel of fun" convention game, I think it was generously described as "unfocused", and we could be a bit more frank than that*!
In a lot of ways, the process for "writing" Doom Town, Horror Victorianorum, Where Zombies Dare and Death on the Streets (1st and 2nd playtests) was to write some characters, write a bit of a situation, then run a play-test and write down what happened. Not exactly rigorous - and not exactly fair on the play-testers.
With Death on the Streets though, I had a break-through after the second play-test. The first playtest was dreadful - truly one of my worse outings. I tried to re-use the timeline format that had worked for HV, but the game lacked the interpersonal spark that gave HV what little life it had. The second play-test was vastly funner, but also pretty rough and with a run-time of almost 5 hours from woe to go, totally unsuitable for the tightly limited game-time at Fright Night.
So for Fright Night, I knew that I'd actually have to legitimately write a game, not improvise then repeat (because I'd tried that, and failed). I looked at every aspect of the game and said to myself, I'd be prepared to chuck it if it'd make the game better. So I sat down with a list of every scene that had been played out at the play-tests. I recorded both play-tests so this was genuinely comprehensive.
I then looked at every single scene and asked myself two questions:
1. What useful fact does this scene convey to the characters?
2. What is a plausible and interesting conflict that can arise organically in this scene?
Any scene that didn't have both was instantly discarded. This left a surprisingly large number of scenes - more than could fit into a session. I dithered about what to cut from where before deciding that I didn't understand the story mechanics well enough to cut anything without risking the game seeming somewhat arbitrary.
And so we played it out. As has been my previous experience with mr_orgue and steve_hix, they played the hell out of their characters, and really pushed their characters' story agendas. They beat the stuffing out of the simple shape of the story that had somehow been preserved through the first two playtests. Which is why I love playing with those guys. :D But! Because I'd prepared every scene explored in the two play-tests, and had really tried to find some point to them, I found myself just re-splicing the scenes out of order.
So at the end of Fright Night, I was left feeling like my notes had won: I'd covered almost everything that had happened in the game in advance - the only thing I had to improvise was one interpersonal investigation, but that was no big deal.
After that, I sat down to write my SDC entry: Spirit of the Tentacle, that got a big write-up on this group-blog not all that long ago. This was a game which had had the full treatment of poor-writing followed by writing-through-improv, and had undergone a similar but not as good transformation as Death on the Streets. However, when I came to look at the actual notes I'd written, they were... not that easy to follow. Gibberish in fact.
So I had effectively to dredge up 10 month old memories of the KapCon run, which went well (and huge props to Norman for basically making that run work despite the remaining flaws.)
The intensively conflict/clue focused write-up I'd done for Death on the streets was just not appropriate. There are no real conflicts in SotT: just pulpy action. So I was forced to revert to my original mode of writing, with the added restriction that it had to be intelligible to third parties.
And in the end, I think I managed that adequately. I wrote it first by effectively copying the high-level story summary that's my usual starting point, then had to unpack each scene in as much detail as I could remember or imagine. To start with, I grabbed a bit of paper, and wrote down how the PCs get into each act, and how they leave it. Those were the crucial things to get right. Perhaps not so much for running the game, but for my own personal development. I focused on those plot transitions obsessively: writing and re-writing them. Then afterwards I went back and wrote the boring location/action details that comprised the meat of each scene.
After the SDC you will be able to judge the results for yourself!
And now I'm writing two games for KapCon 2009. Trail of Blood and Untitled Sophie and Mash Project.
Trail of Blood is in a lot of ways the return to where I began this cycle of KapCon, but trying to do more with your basic horror premise, and using the very cool Dread which won me the 2nd GMing prize at Buckets of Dice. Honestly: with the Jenga tower the game pretty much runs itself in terms of tension and mood.
I began with Trail of Blood by writing a very loose sketch of the journey, and some things which could happen, and am now in the midst of writing the character questionnaires before returning to the game action.
* Without a doubt, my stand-out, absolute favourite CON GMing experience was The Storm Breaks, a game with literally no plot, just four characters that wake up in a medieval/fantasy world instead of the Scottish Highlands. That was a game I wrote a detailed set of locations for, then found myself inspired during the game and improvised a huge amount more detail and new locations, and the players really ran with the immersive character play without stressing about plot: just existing as characters in this weird space. It was pretty randomly good: I doubt I could replicate it as an experience no matter how much I tried.