In this thread, I said:

In another post, I'll write up what I've done so far, and I'll scrawl notes as I go through my ongoing playtest as I learn about the process.  Maybe, as we compare our experiences, we can start extracting some principles.

This is that thread!  This week was the first full week of "serious" playtest for Dirty Secrets, and I'm pretty stoked with how things are going so far.  So this seems like a good time to start this thread.  Besides, Jason has been holding down his end of the bargain, so it's time for me to pony up.

Dirty Secrets is my game of noir detective stories.  The idea had been kicking around in my head for a while, but I started actual development work on Dirty Secrets in August 2006.  The core premise at the time was having to divide attention between actually solving the case and simultaneously resisting the degrading effects on the investigator's emotions.  The initial system was awful.  I never actually playtested it; I could just tell that it lacked that necessary essential spark of fun.  But I kept plugging away at it, twisting this part and bending that part, trying to see if it would settle into place.

In the meantime, I immersed myself in the genre stories.  I watched Chinatown, The Maltese Falcon, The Big Sleep, and Brick.  I read Dashiell Hammett, and then Raymond Chandler, and then (via Ron's recommendation) Ross MacDonald.  I even tracked down some issues of Fell and read them.  Slowly, the style and feel began to settle into my brain.

The breakthrough came sometime toward the end of the year.  I honestly don't remember when I got the idea to use Liars' Dice as a core mechanic.  I know that, in part, I was influenced by the back-and-forth of Dogs in the Vineyard, which I liked.  Plus, it seemed to fit the genre well.  Most conflicts in the genre tend to be between two people, fencing back and forth, trying to discover what secrets the other person has  without giving up his own.  A bluffing game like Liars' Dice seemed perfect.  Plus, with its bidding structure, it has that back-and-forth feel built into the game.

I had some other design constraints as well.  Mostly, I wanted a game that I would actually play.  Increasingly, I have little patience for out-of-game prep.  I want to be able to sit down at the table and get to work.  Prep at the table is fine with me, but I don't want to have to worry about the game in between sessions.  Also, I wanted to make a GMless game.  Again, this is mostly because this is the sort of game that I would like to play.

At the same time, I wanted a game that my wife would play.  Crystal doesn't usually like the free-wheeling storygames, where you aren't specifically associated with a given character.  She actually likes getting behind the eyes of her character and engaging the story in that way.  Thus, I gradually evolved the current setup, which consists of one player who runs the investigator and the remaining players, who play everything else.  I jokingly say that this is a game for one player and many GMs, but there's a chunk of truth in that.

Finally, I wanted a game that combined real player authorship with the ability to enjoy the surprise reveal when the mystery is solved.

I poked at the system for a bit longer, but I soon came to realize that I needed to put the game through its paces.  It wasn't done.  In fact, there were large chunks of the rules that didn't even exist yet.  However, I needed to have a test of concept playtest, partly to reveal areas of weakness in the game, and partly to see if the core concept even worked at all.

So, one night, I sat down with Crystal and Gabrielle.  The prototype was on the table, and we were about to play.  I made all sorts of suicide statements ("This isn't a complete game", "I hope that it isn't awful", and stuff like that), then we got into it.  As expected, there was all kinds of ugliness, but there were three positive outcomes.  First, we proved that the fundamental system concept was functional.  Second, we uncovered all sorts of areas that needed to be addressed.  (For example, how do you handle research in a game where the players just get to make everything up?)

And third, my playtesters said that they had fun.

That was quite possibly the most important item.

Since then, I've worked on refining the design, which has included conversations with various friends.  In particular, Crystal and I had a highly profitable conversation while driving back from Champaign one evening.  Also, I've continued to read detective stories and dig up other movies, asking myself the question, "Could I do this scene in Dirty Secrets?  What about this scene?"

And now we are into full-blown formal playtest.  Each session starts with a report of new rules, then we play, then I gather final feedback for the night, in addition to any comments during the game.  In particular, I try to home in on areas that caused a strong reaction from the players, be it good or bad.  If it's bad, I want to fix it.  If it's good, I want to do more of it.

However, last night, I realized that I also need to be looking for the techniques that we apply during gameplay.  After all, the three of us have gamed together for quite some time, and there are a variety of techniques that we apply without really thinking about it.  If the success of the game relies, even in part, on these techniques, then I need to be sure that I know about it so that I can communicate them in the rules manuscript.  I'm not just testing the rules; I need to observe the emergent behavior that the rules produce and be able to explain to a stranger how to produce similar effects.

I also realized that I've found my answer regarding the age-old question:  "Do you change rules in the middle of a playtest?"  My current answer is, "Do not change rules in the middle of that rule's cycle."  So, you need to see a full iteration of the rule's effects before changing it.  Some rules have a small cycle, so they can be changed fairly quickly.  Others have a long cycle and shouldn't be tweaked until later.  I'm also working with an alpha version.  Once I get to a stable beta, I'll probably tighten up and play through an entire story with the same ruleset.  I'm also not going to be dogmatic about this.  It's working so far, but I reserve the right to change my mind.

So, in summation, some thoughts from my playtesting so far:

--Don't be afraid to run "proof of concept" playtests before your alpha is finished.
--Have friends around who can critique your work and offer helpful insight.
--If your playtest is producing functional play, be watching to see what techniques are being employed so that you can explicitly describe them.
--Be careful when you change your rules.

My current playtest goal is to finish out our current story, which will also be the finish of the Mystery Resolution system cycle, which is my current area of concern.  Once I get that into rough shape, I'll probably write up a formal playtest document.  Until then, it's scrawls in my notebook.

Jason wrote:
Are you maintaining a rule throughout a session, or when you say "cycle" do you mean a single iteration of that rule?

By "cycle" I mean a single iteration of the rule.  However, I prefer not to change the rules in the middle of the game session unless something immediately presents itself as being horribly broken.  Honestly, even in this case, I try to maintain the discipline of not changing the rule until after the session.  After all, something that can appear to be "horribly broken" at the moment may actually be okay when considered rationally.  Sometimes, these items can even be a feature.

I have an example of this.

An Investigation conflict in Dirty Secrets includes an feature called "Violence Dice".  After rolling his dice pool to do Liar's Dice, each player secretly sets his Violence die to any number that he wants.  In addition, a third Violence Die is rolled and placed in the open in the middle of the table.  Once the final bid is called (e.g. "there are 4 threes under the dice cups."), you also resolve Violence Dice.  Each die that matches the die face in the last bid puts one point of Violence into the resolution narration.  1 point of Violence=violence or threats of violence without any long-term injury (e.g. a slap in the face).  2 points of Violence=violence with long-term injury (e.g. being stabbed).  3 points of Violence=death.  Whoever has narration rights must incorporate this Violence into the narration.  Moreover, it must be applied against a Character (someone with a character sheet), and it must be the result of a Character's actions.

This rule has had a bumpy history in playtest.  On the one hand, it certainly produces the occasional shocking violence that I find in the genre.  On the other hand, I've found that players frequently haven't known where to assign the Violence.  I added the rule that it must be applied to Characters to assist in this process and to keep the entire system from becoming a farce.  That has helped, but it's still a tricky item.

So, I guess it shouldn't have surprised me that this rule came up again in playtest last Thursday.  Our investigator is Bob George, an aging DEA desk jockey who is doing a private investigation on the side.  He was tailing Jeff, a burn-out who appears to be involved in the case.  The conflict was over Bob's ability to tail Jeff without being seen.  Bob was ultimately successful, but the narration demanded 1 point of Violence.  Crystal is playing Bob, and she wasn't happy.  The entire point of the conflict was tailing Jeff without being seen.  Where would that Violence narration come from?

Now, at this point, we could have thrown the rule out as broken or needing serious overhaul.  However, I actually decided to stick to it.  Then the light bulb came on.  The rules didn't say that Bob had to provide the Violence.  So Crystal narrated Debbie (Bob's client, in fact), showing up and slapping Jeff.  Then they started making out in the parking lot.

What appeared to be an annoying rules wart turned into another method to drive the story forward.

Of course, while I say this, I also realized that I violated the "watch for a cycle" rule last playtest by tweaking a rule before it had run for a full cycle.  And, honestly, I'm finding that I regret it already.

I find your comments about emergent play and capturing the social dynamic of your group really interesting.  It's certainly a tasty trap to assume that everybody will play like "your guys". 

Very much so.  Of course, at the same time, you want to leave room for each group to find its own equilibrium.

Currently, I'm achieving this with the Challenge rule, which was inspired by Universalis and Spione.  Basically it works like this:  any player's narration may be vetoed by the unanimous agreement of the rest of the players.  This has been working out quite well in play.  What this means is that the group is given authority to police its members without needing a GM, but there is still a great deal of freedom for the individual, since he only needs to persuade one other person that his narration is worthwhile.

Hopefully, this rule will act as an umbrella to allow the gaming group to address the inevitable grey areas that will appear in play.

This actually reminds me of one other design assumption that I'm making.  The ruleset for Dirty Secrets assumes that you have a functional game group composed of players that want to have fun together in the detective genre.  This is a pretty big deal, actually.  What it means is that, from a design perspective, I am ignoring the following groups of people:

--people who are unwilling to cooperate with their fellow players
--people who want to have their fun at someone else's expense
--people who expect the playing of the game to teach them about the genre

The first two cases are easy.  These are problems that simply can't be addressed by the game rules.  If you're at the point where a certain player is narrating nonsense simply because he can, then the Challenge rule won't help you.  Instead, you need to have a conversation about if you even want to continue to play the game.

The last case is a little harder.  Certainly I intend to provide guidance to various detective resources, like a bibliography and filmography.  I may even write a detective short story for the book, although I'm not committed to that at this point.  However, I cannot make someone understand and love the source material, simply by playing the game.  I can't design a game for a player is unwilling to have a basic grounding in the principles of the genre.  That doesn't mean that they need to be freaks about detective fiction.  Honestly, if you have even a passing familiarity with the genre tropes (e.g. the private investigator, the woman who knows too much) then you'll be fine.  But I'm assuming that my players are bringing that with them to the table; I can't give it to them once they are there.

GreatWolf wrote: What it means is that, from a design perspective, I am ignoring the following groups of people:

--people who are unwilling to cooperate with their fellow players
--people who want to have their fun at someone else's expense
--people who expect the playing of the game to teach them about the genre

Right there with you!  I think making a broad assumption that the first two are givens is actually healthy, but it really has to be an intentional decision.  The third, interestingly, I totally understand but worked my way around in my own design (at least I think I have).  But as an example, as somebody who is superhero illiterate, I see this third design assumption in, oh, every superhero game ever written.  Which is fine.  But you'll want to be up front about it, unlike all superhero games ever. 

Jason wrote:
The third, interestingly, I totally understand but worked my way around in my own design (at least I think I have).  But as an example, as somebody who is superhero illiterate, I see this third design assumption in, oh, every superhero game ever written.  Which is fine.  But you'll want to be up front about it, unlike all superhero games ever. 

Very much so.  I haven't completely figured out how I want to do this, but I figure that I'll probably highlight a few basic resources.  My guess is probably "The Big Sleep" by Raymond Chandler, "The Goodbye Look" by Ross MacDonald, and the movie "Brick".  My thinking is that this will help as a guide to those who don't know the genre, plus be an point of connection with those who do.  "This is the Ross MacDonald RPG?  Awesome!"

This being in addition to a bibliography, of course.

GreatWolf wrote:
However, I cannot make someone understand and love the source material, simply by playing the game.

Have you checked out Polaris?  It sort of creates its own genre (doomed fairy knights of a dying people at the North Pole) but does a good enough job of establishing what the setting is about, and what kinds of stories it's prepared to tell, that most players I've seen have been able to grasp it quickly and run far with it.

sn0wshrew wrote:
Have you checked out Polaris?  It sort of creates its own genre (doomed fairy knights of a dying people at the North Pole) but does a good enough job of establishing what the setting is about, and what kinds of stories it's prepared to tell, that most players I've seen have been able to grasp it quickly and run far with it.

Grin!

I am the Platonic form of the Polaris booster.  Indeed, I claim the title of World's Biggest Polaris fan.

Um, in other words, yes, I've checked out Polaris.

However, even with Polaris, there is a need for the players to come to the table understanding the basic genre assumptions.  Polaris helps you by giving you an endpoint:  either you will die or become a demon.  That shapes game play, because players will end up angling for one of these two options.  But, even with that, the players need to grok the Color of the game or it won't work.  Instead, the players will find themselves wandering in circles, not really sure what should happen next.  Check out this thread to see what I mean.

In the same way, Dirty Secrets requires that the players understand both the look-and-feel of the game and the sort of stories that are in the genre before starting to play.  That's what I meant by my statement.