In another thread, Per wrote:

Seth, could you perhaps share your thoughts on how the initial characters in the game are written - why did you choose this approach, and why are characters created during the game done differently?

Okay, I have a little time, so let's jump into this.

I will cop to being a selfish game designer.  I really only want to design games that I want to play and will fit into my life.  So, one of my personal bugaboos about roleplaying is prep.  Frankly, I hate it.  I don't really have the time for in-depth prep anymore, and, honestly, I don't want a heap of homework to do so that I can enjoy the next game night.  I got into this hobby to play and be creative, not do spreadsheets!  Yes, that's an unfair characterization, but it was fun to write.

So, one of my ongoing design goals is to reduce or eliminate prep from my games.  Legends of Alyria took the first step, with its storymapping and initial prep session being built into gameplay.  So, rather than the Narrator having to prepare something, the group preps for it at the table.  These days, everyone points at Primetime Adventures and thinks, "So what?"  But when I originally wrote Legends of Alyria, it was a fairly radical idea.

So I decided to take the next step with Dirty Secrets.  I've been playing a lot of German boardgames over the last five years, and I've wanted a roleplaying game that was as easy to set up as a boardgame.  Just think:  most boardgames require no prep.  You just pull it off the shelf, sit down, and play!  That's what I wanted out of my roleplaying, and that's what I wanted out of Dirty Secrets.  Also, I knew that I wanted Dirty Secrets to be a player-driven game, while still having the shock of the Big Reveal.  So this was another reason that GM-based prep was out.  Finally, I wanted Dirty Secrets to be a GMless game, because I like playing GMless games.

Therefore, I needed a system that quickly generated a starting situation to allow the gaming group to get past all that annoying prep and into the interesting bit, which is actually playing the game.

I don't know exactly where I got the "pass stuff around the table" idea.  At the very least, I was probably thinking about Burden creation in carry, but there may have been other games in the mix as well.  Anyways, this method allowed for a semi-random generation of characters, which I liked.  Another feature is that, once you label the appropriate Character cards as being investigator, victim, or suspect, the players have an opportunity to shape the starting situation without having complete control over any one element.  As a result, the starting Characters are nothing that any one player would have chosen to do.

Also, it breaks down the decision matrix.  Rather than having to dream up an entire character, each decision point was fairly simple:  choose a Demographic, and then select an option from the list.  So, if you had the investigator sitting in front of you, you could think, "It would be neat if the investigator is just a regular person" and therefore select "citizen" for Legal Status.

Towards the end of development, I added the veto mechanic.  If you look at it closely, it's really just the Appeal mechanic, but turned inside out a bit.  This was to solve a specific design problem:  how can you let a player Appeal part of a character, if no one can definitely be said to be the original narrator?  I also wanted to introduce Appeal here to prevent too much silliness in the initial situation.  This turned up a lot with senior citizens.  The rules have explicit limits on minors (which is the only explicit concession that the game gives to Lines, by the way), but the game says nothing about senior citizens.  So playtests were constantly turning up Characters who were 80 or 90 years old.  I couldn't Appeal, in that rules draft, so I added the veto rules to give players some control after they saw what the process had wrought.

(At this point, I want to publicly thank Paul Czege for being the person who finally pushed me over the edge to institute a rule to let me stem the tide of senior citizens in my games.)

This went back to a basic design principle that I had established for the game.  I didn't want the players to be able to override the objective decisions of the game mechanics.  This meant that I needed to be careful about which decisions I handed over to the game system, but I also wanted to make sure that any decisions that the game system made were irrevocable.  So, for example, this is why you select a Crime to be resolved after seeing who did it.  Originally, you picked the Crime first.

So, those were my thoughts about setup.

During the game is completely different.  Technically, creating a Character in the game is a two-step process.  First, someone narrates the existence of a person.  Then someone declares that person to be a Character.  They don’t even need to be the same person.  However, when a player declares a Character, he is pointing at a person that already exists.  By this point in narration, some details have already been established, like Sex or Age or sometimes Race.  To “pass the card” at this point doesn’t make any sense.  Instead, the Authority is essentially writing down what has already been established in play.  At this point, players can Appeal the Demographics if there’s a disagreement about the details that were already established, because it’s now clearly the Authority’s choice to write down a given Demographic.

The distinction between person and Character is important.  As I point out in the book, “By [declaring a Character] you are telling the rest of the players in the group that you want this person to be important to the story and that you are committing to make this Character interesting to the rest of the group as well.”  (p. 82)  So, a “Character” is a special subset of all the people that will pass through your story.

Which people should be Characters isn’t always apparent.  This group had a person who was a fairly visible part of the story but wasn’t declared to be a Character until almost the end of the game.  Here’s what he says:

There's one scene in particular I would like to talk about. The opening scene with Lucy (who was at this point a non-character) and the investigator. In it, Lucy describes the murder victim and how she was murdered, going into a fair amount of detail, including a blatant rip-off of, or, if one is generous, an homage to, Chinatown, all of which sets up a lot of the action to come.

It was a great opening scene in terms of serving the purpose of opening scenes (laying lots of clues to be followed up on in the middle part of the story).

Later, Lucy is now a [C]haracter, and it turns out that the murder victim is actually a suicide victim and that Lucy helped cover it up. But, in the opening scene Lucy said it was a murder victim, and spoke about the murder in a very CSI manner, giving out lots of details.

Meaning, in the opening scene, Lucy was lying through her teeth. But the really amazing part is, in that opening scene, the two advisors, the investigator, the investigator's player, and the narrator, all thought she was telling the truth.

Now the opening scene is incredible because it not only set up action, but it also had a character who was deeply involved in everything, up to and including the investigator's back story, lying big time and nobody knew it.

Totally, Seth, thanks for the comprehensive answer :)

Cheers,
Per