Daniel Cook wrote about feedback loops, learning and game mechanics in his blog Lost Garden a couple of years ago (he then wrote another article about this on Gamasutra). What has stuck in my mind about these articles is what DanC describes as skill atoms. Skill atoms are things that can be learned by playing the game:

• The player interacts through the game, does something,
• the game responds to the action,
• the game provides feedback on the results
• player interprets the feedback and essentially creates a model of how the game works in similar situations.

The player learns a singular skill by looping through this sequence multiple times. This is called a skill atom. An example from DanC's article is from a Mario Bros-like platformer game.

Learning successive skill atoms usually requires the player to have mastered some of the more simple skill atoms. The example (this one too is from DanC's article) is also from a platformer.

  DanC suggests improving the method of game design by designing skill chains out of these skill atoms and designing the game mechanics and levels based on those. This sounds like tremendous advice for designers of video games.

I recently got thinking that the same kind of feedback loop, only possibly more complicated, happens also in role-playing games. I got to thinking about this as I was trying to come up with an article on how to get the most of the excellent modern war role-playing game Shell Shock.

We started a Dogs in the Vineyard game a couple of weeks ago with a character creation session.

This was the first time I played RPGs since Senja was born.

Matti (being the GM) explained the game briefly and then we went through the character creation process. By my suggestion we then proceeded to talk a little bit more about the game and our expectations before creating the characters. We read bits of text from the rulebook and the introductory one-page pdf and we talked about the tone and style of the game we would like to play.

After that we started the actual process of character creation. We had agreed that we all wanted a colorful group of Dogs with some variety and difference of opinion in it. We ended up choosing different backgrounds for each one. Each player bounced his ideas of character concepts aloud at first. Then we kind of went through informal question rounds of 1) what kind of character I'm thinking about and 2) what kind of Dog this character is (each one adding a different viewpoint for his character's dogness and saying how it differs to the others characters).

After that we started to fill out the sheets. This was also quite collaborative as we asked for second opinions on stat allocations and trait naming quite a lot. After we had filled out the details each of us presented his character briefly.

The final part of the evening was playing out the initiation conflicts of the characters. That way we got to try out the conflict system in action. What we found out that the initiationconflicts were pretty tough; I think no-one won his conflict. The other thing we figured out was that losing 'just talking' conflicts was in fact mostly advantageous as the character got experience but there was really no danger of major fallout. We immediately decided to look for conflicts to lose in our game :)

Our group of Dogs turned out pretty interesting. There is one half-mountain folk dog, a sullen goth-type dog with a great legacy and no social skills to fill the boots of her mother (!), a former street urchin with a lot of repressed hate and one quarter-back type youngster that is ready to make the world a better place for everyone (that's my character).  

We've now played one session of the game; I hope to be able to write a little about the actual play later.