Shooting the Sacred Cow

[Valamir]

I was going to post this in Indie Design, but then I decided to talk more in generalities than about the specific application to my on going Robots & Rapiers project so I figured this is a better slot for it.

First a bit of background.  The Alpha test rules came out for Robots & Rapiers back in September, and by November I'd had sufficient play test feedback to move ahead and revise the rules to a beta level.  My initial naive belief what that I'd have the beta update ready by Christmas in the vain hope that someone would be looking for something to play over the holidays.

Needless to say, it didn't happen.  What the Alpha revealed was exactly what a good Alpha should.  Conceptual and structural issues (balance tweaking stuff is more of a late Beta concern).

Alpha Verdict:  Great start, but the game just wasn't playing the way I wanted it to.  Actual play did not mirror the image of actual play that I had in my head.

I figured, no problem.  A couple of brain dump sessions with some of my playtesters and I figured that all I needed to do was make a few adjustments.

Boy was I wrong.

See, my initial diagnostic and ideas for adjustments had just looked at the symptoms.  I hadn't yet encountered the real problem.  The real problem turned out to be TOO MANY LAYERS.  Stats, substats, modifiers, etc...figuring out what to roll for a simple check just involved too much.  Not too much calculations, just too much remembering all of the different layers that were involved.  Plus there were some holes in the layers so that some things you might want to do weren't immediately obvious how to go about combining the various layers.

Weeks of rewriting later, I'd fixed it...until I took the highly advisable step of imaginary game play.  Not dreaming about the ideal, but imagining actually sitting at the table and telling players what they needed to do to make a roll to accomplish X.  Turns out my brilliant solution to the problem of having to many layers...was to add more layers...ACK!

The game was spiraling out of control into some monstrosity that even *I* didn't want to play.

What to do.  Back to basics right.  Strip out the chrome and start over.  

Right.  Except for one problem.  Sometimes, that isn't enough.

Everyones probably familiar with the old saw  "Better, Cheaper, Faster;  pick two" (in fact a fun little game mechanic was based on that idea awhile back).  

That was my problem.  I had 3 things I wanted to accomplish but after several different versions and thousands and thousands of words over the next several weeks, I couldn't figure out a way to get more than 2 of them at a time.

For R&R, the three things I wanted to do were:

1) use traditional based mechanics and game design structure (i.e. no wacky Inspectres, Universalis, the Pool type stuff).

2) Build a system that could reflect through changes on the character sheet the transformation of the character from a programmed robotic slave of the "Tapestry" to a free thinking independent sentient being.

3) Fast Fast Fast action oriented play where whipping up feats of derring-do on the fly and getting to the result could happen ...well...fast.


The layers problem was what was, of course, killing #3.   But when I got rid of the layers I wound up losing either #2 or #1.  This was the dark hour of the project indeed.  It actually really negatively impacted my enjoyment of the holidays because it was hanging over my head like a dark cloud.  I was ready to just scrap the whole project, or just release it as a Donjon or Sorcerer mini sup (which had been my initial plan anyway)

Aside:  I think pretty much EVERY game project goes through this dark stage.  At least every GOOD project.  I imagine projects where the designer doesn't really give a rip means he never really stresses over whether it works or not.  My advice, if you're REALLY serious about getting a game to publication...spend money.  However much money it takes so that you'd be feeling it if you lost it.  Don't rely on that initial enthusiam to get you to completion (it won't, which is why every gamer on the planet has half a dozen half finished designs on their computer or in notebooks).  I can safely say, that Universalis would never have been finished if I hadn't gone many hundreds of dollars in the hole that would have been pissed away if I didn't finish and sell the thing.  Pretty grim but it works.  And it worked for R&R too.  God bless paying artists in advance.


For most of January I struggled and wrestled with these issues to no avail.  Every night I'd write a half dozen pages, and every day I be crushed by how bad they sucked.  Gradually, a pattern emerged.  All of those rewrites...the long drives thinking about what was wrong...paid off.  I could now see the monkey wrench in the works.  The obstacles that were causeing all of the problems.

To put it generically, Rule X sucked.  But I needed Rule X (or something close) because without it design parameter A couldn't be accomplished.

BAM.  BOLT FROM HEAVEN.  ZOWIE.

It wasn't Rule X that sucked.  It was design parameter A that sucked.

God Damn, how did I miss that!   Parameter A isn't even a design goal...its a presumption about how best to meet the real design goals (the three items above.  It (along with Parameters B, C and D) were the obstacles that all of my layers were attempting to work around.

That's when it started.  The shooting of the Sacred Cows.

You see most of those parameters had been with R&R since the very beginning of the concept.  One of them was in fact an early eureka moment.  It was the design idea that had convinced me that making a game like R&R was even possible.  That idea was what got me started thinking I good design a stand alone game rather than a Donjon supplement because I figured out a way to get what I wanted that I couldn't figure out how to do in Donjon.

Yup...but now it had outlived its usefullness.  I'll always be greatful to that idea for giving me the courage to head down the road...but now it was long past time to part company with it.  Eureka moment of profound epiphany it may have been but now it was an Albatross.


I'm writing in generalities here so as not to bore you all with the painful details of the R&R design specifics.  But alot of these Sacred Cows are things that you probably have in your game designs and which you may find your design gets better from a little culling of the herd.

Some Sacred Cow examples.

1) Attribute names.  Yes Attribute names.  These are insidious little silly fuckers.  During some brainstorm you come up with a clever idea for an attribute.  Maybe its a unique attribute that measures something no other game measures.  Or maybe its a favorite attribute you found so useful in another game that you once promised yourself you'd include in any game you designed.  Maybe its just a really kewl sounding name "oh yeah, I got have an attribute called X...that rocks".

You then to all kinds of design gymnastics to try and figure out how to incorporate this thing into the design.  You may even be convinced that it fits into the design well.  You MAY even be convinced that it is the lynchpin, the cornerstone of the whole design that makes everything work.

Maybe it is.  But maybe its just a sacred cow that needs to be shot so you can get back to designing the REAL cornerstone of the design.


2) Sacred subsystems.  During an early version of the rules you came up with a really cool way to handle character wealth, or character advancement, or reputation, or contacts, or whatever.  Its brilliant.  Its the coolest way of doing this you've ever seen.  Then the version of your main rules moves on and leaves the subsystem behind.

Wait a minute, the subsystem, no longer works the way it used to because this or that piece of the main system has since been changed.  That sux!  This subsystem was your proudest creation to date.  You've bragged about it on all the web sites "well the way I handle that in ==super cool game== is...".  So you twist, you turn, you backpedal, you rework the main rules so that the subsystem you're so proud of is up and working again.

Thing is your main rules worked just great.  They were outstanding in fact.  Until you started to futz with them to make them compatable with that favorite subsystem.  Now you have a really brilliant idea stuck in a shitty game.

Time to shoot the sacred subsystem dead.  Maybe you can figure out a variant way of using it in the game.  Maybe you save the idea for a future project.  Whatever you do with it, you don't let it ruin your entire design, no matter how brilliant it seems.


3) Basic Simulation assumptions.  You have a design goal.  You want to design a game that accurate captures the flavor of X.  You have a great idea on how to do that.  A mechanic that works like ==this== (a die pool, a resource that gets spent, a way to generate modifiers, whatever) will really capture the flavor of X that I want.

And so it goes.  That assumption about that mechanic begins to underlie every design choice you make.  Hey that's a good thing, right?  After all "system matters" right?  Yup, right up to the point where the game stops being about the design goal, and starts being about supporting that mechanic.   Buzzzz.  Now you're on the wrong track buddy, and its a long hard road back, because that damn mechanic has its tentacles into everything you've written.  Your game no longer really accomplishes its design goal, but its too late to go back and change everything now...

Nope...its never too late to shoot the sacred cow.

Sometimes, you just have to go back to formula...all the way back.  You may find some salvageable pieces along the way...but when its time to start over its best to start with a blank slate and then discover what of the old you can add back in, rather than start with the old and try to cut the bad parts out.

Sometimes you just have to make a nice big pile of ground chuck out of your assumptions.

[Matt Wilson]

Holy Sacred Cow, Ralph, I have been and still am dealing with all that same stuff.

I started working on this game I had an idea for that was all based on a setting concept, and at one point I noticed that I had drifted a lot from my original idea, and figured I should focus. So I broke a bunch of rules loose and started building PTA.

What I realized over a few playtests was that the conflict system, which I had dragged over from that other game, offered nothing that facilitated the kind of play I wanted for PTA. Whoops.

So now it's long gone, but man it was hard to come to terms with that. The mechanic was the place I started, and I imagined it being used in so many cool ways.

[ethan_greer]

Preach it, brother!  I 100% agree with everything you're saying, Ralph. Great post.

[Matt Snyder]

Ralph, I encountered almost exactly these same quagmires in designing Nine Worlds. You can go on the Chimera Creative forum, re-read posts (many of which you were involved with) and watch me struggle through it.

I read through those posts last night, and boy was I embarassed at how much of a jackass I was to people offering insight. (Specifically, to Mike Holmes, you, and Ron Edwards.)

But, I also realized that I wasn't wholly wrong in my incomprehensible flailing about. I just didn't know how to assess the problem. Re-reading those only allowed me to re-assess the problem last night. It only took, what, 6 friggin' months?!? Sheesh.

So, yeah, I'm with you, too. You may wish to go back and read some development on the forums or in your own notes. It may astound you how "alien" your own comments may seem, even those from several weeks ago. I think it could help you that much more in solidifying your core design goals.

[Jonathan Walton]

God bless paying artists in advance.

And the people say "AMEN!"

Honestly, spending $600+ on art is the only thing that's going to force me to finish and publish both Argonauts and Facedance/Beneath this Facade (how many frickin' titles will that game have before it's through?).  This tactic is a great one.

I also agree with shooting cows.  FD/BTF really, really really needs me to go out back with the shotgun and blow holes in my darlings.  I just haven't gotten up the guts to do it yet.  But, in the long run, it'll be a much better game for it.

You post was rather inspirational, I must say.  Makes me want to start loading both barrels...  I'll probably bookmark it, so I can come re-read it whenever I get squeemish about axing something beautiful.

[Valamir]

Heh, no kidding.

You may be pleased (and perhaps a bit annoyed) to know, Matt, that I had 9 Worlds in mind when I started pulling the trigger.  It suddenly dawned on me that my conversations with myself about R&R were sounding alot like...well...like your conversations abut 9W (and alot like my conversations with Mike about Universalis for that matter).  

I was all like "you talk the talk when its Matt's game.  You start to wonder if he's going to step up and do what needs to be done with it or if he's going to let the idea die...time to shut your mouth and ante up boyo"

Bang Bang.


And yeah, I've gone back and read through the old Uni threads now locked and buried in the forum.  Damn...that game sucked...

[Paganini]

There's a saying in the writing industry to the effect that, once you've identified your favorite part in your work, that's the part you should cut out without thinking twice.

This applies pretty well to RPG design. I shot a bunch of Draconum's sacred cows last night. If something bugs you about a game, usually the first response is to try and fix it. A lot of game design happens as a result of a gamer not liking some particular subsystem of D&D, and trying to write a game that fixes it - straight up Fantasy Heartbreakers. I've learned that the first question you should ask yourself is not "how do I fix it," but, "is there any reason for this to be here in the first place?"

[Jack Spencer Jr]

Here, here, Ralph. Sir Arthur Quiller-Couch once said, "Murder your darlings," and he was right.

That is, if I understand you correctly in that you're not talking about sacred cows, which would be darlings for the mass populous of roleplaing culture, but darlings for you, the designer, as you're designing your own game.

[Doctor Xero]

Painfully true.

Although sometimes one can take the system out and use it to design a
new system (sort of like splitting in blackjack).  This is done in writing --
remove a beloved section, write the story without it, use that section much
later as the seed for a new story.  Sometimes the same rejected sacred
cow can sire several good stories before it actually results in one in which
it fits!

Doctor Xero

[C. Edwards]

Ahhh, if only this thread had been started three months ago. I took a whole gang of my little darlings, and a Sacred Cow or two, out behind the shed a couple nights ago and blew their heads right off.

It's not too hard a thing to do once you realize they were just getting close to you so that they could whisper little lies in your ear while you sleep.

Anyway, I'm right with you, and boy what a relief it is.

-Chris

[clehrich]

Apart from saying "amen" to all that and more, I'd like to add a comment from the non-fiction writers' guild forum of academic crazies [read: I teach and do academic writing, although I struggle with RPG writing].

Ralph beautifully lays out 2 of the big ways to deal with a grotesquely huge problem, which is to say the ways to spot the actual pea under the bed that's causing the pain.  

Okay, it's a mixed metaphor, we don't do metaphor over in the academy, bite me.

1. You started with a great idea, and have fleshed out everything else, and it may be that the idea you started with, however cool, just ain't so.  This happens constantly in research of whatever kind (although I don't know personally with the sciences): you have a neat idea, which drives you into researching something, and that drives you, and so forth.  Eventually, you end up with a spectacular article, but there's this nagging problem somewhere.  Chances are, it's one of those things that you thought was cool and interesting at the start.

2. You know there's a problem, and everyone tells you there is a problem, and you pound at it and pound at it until BAM, bolt-of-lightning time.

Okay, but there are two others.  For practical reasons, they don't always help.  But....

3. Outline the thing as essentially a mathematical proof.  You have to argue that every bit requires every other bit, rigorously.  You do this in bits, in tags; you don't write it out in sentences.  You do it structured as an outline, so it's all visible.  Suddenly you spot something, because to do this you are effectively outlining what you have written: in this place and this place, I say that these two bits support each other.  But they don't.  I now only sort of remember what I was thinking, but these days, after all those revisions, I know for damn sure that the final version doesn't work that way.  Now that I cut away the support from one of these bits, I see that it has no connection or support anywhere else.

4. You're not going to like this, but this is the #1 best way to edit and revise your own work:  drop it for 6 months to a year and do something else.  And by do something else, I mean write something else.  Learn on the same "channel" as what you're doing now.  When you come back, you're going to think, "God, was I such an asshole that I actually thought that was a good idea?  What a moron!  Good thing I didn't try to publish that piece of moose-turd-pie."  I know, your supporters may be banging down the door.  In the academy, it's the people on the tenure committee.  But honestly, 6 months of doing a different project of a similar type will solve a disturbingly large number of problems, because you stop being in love with poor old Bossy and can send her to the knacker man.

Just my $.02, as someone who (if I say so myself) writes quite well academically, but can't seem ever to get his enormous Holstein of a game cut into tasty burgers.

Chris Lehrich

[Rob MacDougall]

Just my $.02, as someone who (if I say so myself) writes quite well academically, but can't seem ever to get his enormous Holstein of a game cut into tasty burgers.

Nicely put, Chris, and good advice. Don't wait too long, though. While I'm not on your tenure committee, I am at least one member of the academic crazy guild who is very eager to see your game make it to all-beef patty form.

[Valamir]

Hey Chris,

Yeah, #4 absolutely is a good one.  For me it doesn't usually take 6 months of writing something else.  For Universalis it was play a single huge massive 200 hour game of Civ III and actually play to the very very end, which took about 2-3 months.  Actually I suspect that rather than writing something else, the best "palate cleansing" technique for me is doing something completely unrelated to writing in any form.  For me that tends to be any of a variety of conquer the world and slaughter my enemies games...there's probably a psychological diagnosis in there somewhere.

Tell me more about #3.  I'm not visually groking the structure of such an outline.

[clehrich]

For me it doesn't usually take 6 months of writing something else.  For Universalis it was play a single huge massive 200 hour game of Civ III and actually play to the very very end, which took about 2-3 months.

Good point, Ralph.  My friends (and my wife) can't understand how it's possible to do serious academic work and at the same time do RPG stuff and also play video games.

Actually I suspect that rather than writing something else, the best "palate cleansing" technique for me is doing something completely unrelated to writing in any form.  For me that tends to be any of a variety of conquer the world and slaughter my enemies games...there's probably a psychological diagnosis in there somewhere.

Yes, but I want to stress that just "taking a break" doesn't cut it.  You actually have to get the thing far enough out of your mind that you actually forget a little bit of what you've done.

I'd compare it to taking a high school math test.  If you're pretty good at algebra or whatever, and you go through the test fast, the teachers always tell you to "go back and check your work."  But most of the time, you can't do this effectively, because you still remember the problems and so you still remember why you did what you did.  The same happens with any kind of writing, and you need to get far enough away from it that you actually don't remember exactly why you chose what you chose, why you did what you did.  I find that writing something else is very efficient, because that part of your brain sort of gets filled up with the something else, but I can see that destroying the world could work too.

Outlining as Revising--this is quite long

Tell me more about #3.  I'm not visually groking the structure of such an outline.

Yeah, my students don't get this either until they do it a bunch of times, but I'll give explaining it a shot.  It's a little tricky to see this with RPGs, so start by imagining an argument essay, i.e. one where you have to argue that you're right about X and so-and-so is wrong about X.  What you do is to make an outline (as in with the multiple indentations and whatnot), like this:

1. Every paragraph gets a heading.
2. In sections that seem problematic, every sentence gets a heading.
3. Every cohesive block of paragraphs gets a heading.

It's easiest to go in this order, actually, because you can simply play Jeopardy with the paragraphs: "to what question is this paragraph the answer?"  If you can't work it out pretty damn fast, you need to do step #2; in addition, if you can work it out very fast and it's simple, but the paragraph is very long, then you need to do step #2.  Once you've got all the paragraph headings, you can also spot the big blocks (step #3, although half the time you wrote this as a block anyway, obviating the step to some degree).

Now you have to understand that each "entry" here is on the order of 5 words.  Don't bother with articles and conjunctions and whatnot; just get a kind of bullet-concept version.  Remember, it's for you -- nobody else has to be able to understand it.

Here's the point:

1. When you break down a paragraph this way, you will spot the following things happening

[*]same thing said 3 different ways
[*]logical order unclear or confused
[*]missing logical or other step
[*]whole paragraph could be said in a sentence
[*]one point or sentence doesn't belong in this paragraph at all[/list:u]2. If you break down a sentence this way, actually, you can get a very graceful sentence out of a horror, but I'm not going to worry about the sentence level here.

3. When you look at the whole block of text this way, you see the same things as at the paragraph level, but they mean different things.  For example, you may notice that you explained X before Y, but that if it were the other way around it would take less explaining.

4. All revision of structure should now take place on the outline.  Stay away from the drafted text.  Just shift things around on the outline until it makes good sense.  Try, insofar as is possible, not to think about what you actually wrote, and simply keep it in little block units like you've got on the outline.

5. The final reordering should be something that you can explain to someone else, in this order, following these steps.  Here are some ways to know that you have succeeded:

[*]several paragraphs have been shattered into component parts and distributed across large blocks
[*]several paragraphs have been made into single sentences
[*]some blocks have been shattered and folded together
[*]any introductory and concluding material now makes little sense with the rest of the outline, and will need to be totally redone
[*]where there is a section that has progressive steps, as in an argument or a list, you cannot swap the order of steps without expanding the number of steps
[*]overall, the whole outline is considerably shorter than it was
[*]there are probably fewer levels called for in the outline, or at least there are more entries at higher levels and fewer at lower ones[/list:u]Now go and cut-and-paste the actual document to fit the outline.  Create a cutfile of the stuff that no longer appears in the outline.  Print it all out, and set to work with ye olde blue pencil.

Principles of this last revision:

[*]Never add to the outline
[*]Never, never use 2 where the outline says 1.  That is, don't write 2 paragraphs if the outline calls for 1, and most of all don't write 2 sentences if the outline says 1.  If you can say it to yourself in 5 words, you can say it to a reader in one clean sentence.  This is the most important and hardest part about this whole process.
[*]Construct sentences and paragraphs in order to do what the outline says, and don't allow tangents
[*]If you have something you loved from the first draft that now has to be cut according to your outline, put it in the cutfile and wait
[*]Don't do examples, back-story, or other frills now, as they will only distract you
[*]Do the introduction and conclusion, which definitely need a total rewriting, last and from scratch[/list:u]Finally, you have a total revision that at least makes logical sense.  Now you need the examples, and chances are a lot of the ones you already constructed last time (and which are now in the cutfile) can be inserted without much change.

After all this detail, what's the actual point?

By abstracting your work to such a degree that it's almost unrecognizable as prose or even as your work, you get the distance required to revise.

Anyway, that's my little manifesto of outlining as a revision technique.  

My students hate this, and I make them do it quite a lot.  By the end of a semester, however, all the better students (i.e. the ones who actually did what they were told and didn't weasel around it) discover that they can actually write most of their papers this way, revise in outline form, and only actually set it all down in real prose near the end.  That way they get one really sophisticated draft, rather than the slow agony of a rough moving gradually to a final.

For anyone who actually cares in the long term about writing, by which I mean anyone who isn't only concerned about the grade at the end of the semester, applying this method in combination with the "lots of time away" method and everything else is very effective.  It provides an actual method for the "tear it all down and rebuild" that doesn't involve simply generating a new version with its own completely new problems and half the problems of the old version.

This method works amazingly well.  It is, however, tedious and annoying.

Chris Lehrich

[Valamir]

Sharp.  I wish I'd had you as a teacher for one of the many writing classes I took.  I've stubbornly refused to outline since the day my junior high english teacher first asked for one and now I find it very difficult to conceptually think in outline form.

Instead I conceptually think in Bold Section Heading form which is hierarchally the same as outline headings but much less neat and concise.  In other words, I do what you do, but I do it with entire blocks of text rather than 5 words nuggets.  Thinking in 5 word nuggets is a lot harder to do than it is to to write about doing it...when one isn't practiced in it anyway.

I know that alot of the bullet points you note are problems that I have in my writing (saying the same thing 3 different ways in 5 different places) and for me it means endless rewriting at the draft level to purge it all.

Any good texts you know of that tackle this notion in some depth.  Might be advantageous of me to try to learn some new tricks if I'm going to be writing alot in the future?