At GenCon, Thomas Robertson, Shreyas Sampat, and I (along with sideline commentary from Selene Tan), playtested this game concept that I had. The game itself was fun, but, more importantly, in the process of play we discovered something that I suspect might be one of the new, new things in design mechanics: "key chains." They are so awesome that I've really wished they existed in every game I've played since GenCon, because they were an excellent tool for:

- solving the "okay, but what do we do?" problem
- defining characters in a non-quantitative way (i.e. not using numbers)
- character development
- making the character sheet dynamic and not static
- pacing (Shreyas: "I spent like ALL of the next day talking up how cool that pacing thing was")
- scene framing
- creating story arcs
- defining "episodes" of play

I'm gonna try to explain how they work and why they are awesome.

"Key chains" are named after the Keys in Clinton R. Nixon's The Shadow of Yesterday. You can read about Keys here. Many people have noticed that Keys are an excellent way to track changing character goals and reward players for persuing them. They define what a character wants, but they can change over time based on the circumstances a character finds themselves in. They also leave most of these choices in the hands of individual players. Keys additionally serve as springboards to get the action going, because you know what a character's central issues are. Clinton gives credit to Sorcerer and The Riddle of Steel for informing his creation of Keys.

The original idea behind Key Chains was to have every significant character decision or accomplishment be a fulfilment of a Key, improvizing Keys and stringing them together as needed until they formed a record of play and a record of character development. These improvized Keys would be very minimalist, simply a statement of the character's current goal or accomplishment, which would change over time. There would be no XP or other mechanics attached to it. These Keys would them be linked together in various thematic Chains, showing the character's advancement in various "story arcs" or "character arcs."

I've encountered the idea of "arcs" mostly in writings about comics and film. The idea is that a character or story has various plots and subplots going on at once. A narrative chunk of progress along one of these plotlines is an "arc." For example, let's invent a Buffy episode. The main story arc look something like:

Heroes Notice Weirdness
Heroes React to Weirdness
Heroes Investigate Weirdness
Heroes Determine Cause of Weirdness
Heroes Stop Weirdness

While this is going on, Buffy might be experience her own character arc, which goes:

Buffy Meets New Boy
Buffy Flirts With New Boy
Hints of New Boy Being Evil or Somehow Wrong
Buffy Ignores Hints
Buffy Dates New Boy
Everything Goes Well For a While
New Boy's Dark Secret Revealed

Now, Buffy's character arc will probably take several episodes to complete, while the central story arc will probably be run through in a single episode. So a system to model this needs to be able to track multiple arcs across a variable number of sessions, and let them resolve at different rates. Also, in roleplaying, you don't necessarily know where a given arc is going to go, because the path of character and story development is generally negotiated by the players during the course of play. What other players choose to do affects the next step you able to take in completing various character and story arcs.

[Note: this is all based on structuralist narratology, by the way. The granddaddy of this field is probably Vladamir Propp (1890-1970), who wrote Morphology of the Folktale in 1928.]

Now, there are lots of ways you could potentially set up Key Chains. I'm fascinated by several distinct methods, but I'll just talk about one of them now.

1. Declare and Record a Goal
2. Frame a Scene Where You Make Progress Towards that Goal
3. Record the Result
4. Repeat Steps 2-3 Until You Are Satisfied with the Result
5. The Result Is Most Often Not Your Original Goal

In the GenCon playtest, we just took turns between players, letting each one frame a scene based on one of their Goals.

Shreyas's character Kusa, for example, decided that he wanted a new "Relationship" trait. He declared and recorded the goal "Find New Teacher." His progress along this character arc went as follows:

GOAL: Find New Teacher
SCENE 1: I'm Outta Here (got into a fight with his old teacher and sailed off to look for another)
SCENE 2: Caught in a Storm (in the midst of sailing, ran into a huge storm)
SCENE 3: Kayakwrecked (his kayak is flipped and tossed by the waves, he ends up lost at sea)
SCENE 4: Try Again (after being rescued by pirates, and working on some other arcs, he leaves the pirates to keep searching)

This is as far as we got. Notice how Shreyas is now four scenes into working towards this Goal and really hasn't made much progress. This might change soon or it might not, depending on how the game goes. Eventually, he may decide to go with a result vastly different from his original goal, in which case he'd make up a new trait (not necessarily a "Relationship" one) based on what's happened to the character.

New idea that just struck me.

NPCs as Key Chains. They develop as characters find out more information about them. A great way to have stuff be determined in play and only create as much information as you need. The more players pursue certain NPCs or setting elements, the more complex and interesting they become.

Jonathan: Is the disconnect from mechanics (no XP, no other mechanic reward) a desired feature? Or is it just the current state of the art?

Because right now, I'm not seeing the qualitative difference between Key Chains and (for instance) a character diary. Both make you conscious of where you are, and then you use that orientation to help roleplaying.

But my experience with actual TSOY Keys is that they go well above and beyond that. They inform your actions through the powerful mechanics of reward. So, in your Shreyas example, I'd be more excited to see something like:

I'm Outta Here
1 XP: When you doubt your superior's wisdom internally
3 XP: When you explicitly rebel against good advice or an order from legitimate authority
5 XP: When you lie about obeying, attempt to disobey in secret, and are caught.

Buy-off: Leave your superior's care and strike out on your own, or humble yourself and submit to his authority

I look at that and I think "Wow, that excites ideas for playing out that phase that I wouldn't have thought of. Like, I want to hang around a while so that I can explicitly rebel ... at least three times, before it gets old and I decide to buy off the key and do something else with the points. And I probably want to sneak off and do something I've been forbidden to do, like challenge a local martial arts master and get my ass handed to me for 5XP, at least once."

Does this sound like something in keeping with the direction you're going, or am I striking out in a different direction?

I think Shreyas is right here. The problem with mechanical constraints is the problem with all constraints. They're constraining (duh).

We talk a lot about how good constraints can push us to be creative, and it's true. But at the same time strong constraints reduce our options (which is how they push is to be creative).

So, using Tony's above example, what if I think that there's a much cooler thematic thing to be said by lying about obeeying and not being caught? Sure I can still do it, but the mechanics actually discourage me from it because they explicitly reward me for taking the opposite path.

That's not to say that mechanical constraints are bad, far from it. But it is to say that they're far from a perfect solution. I guess that lacking a better term for it, mechanical constraints like TSOY Keys are 'harder' than freeform stuff like the game we played at Gen Con.

(I also feel that in the interests of disclosure I should note that I struggled with the very same issue while we were playing. I felt as if the chains we were forming were to 'soft'. Where was the mechanical bite? It was only after the game, when I looked back on the experience, that I started to realize just how well things worked out precisely because the mechanics didn't have to anticipate where we wanted to take the story.)

Thomas

Posted By: Thomas RobertsonSo, using Tony's above example, what if I think that there's a much cooler thematic thing to be said by lying about obeeying andnotbeing caught? Sure I can still do it, but the mechanics actually discourage me from it because they explicitly reward me for taking the opposite path.

But ... then ... you'd have built a different Key, wouldn't you? One where you get rewarded for disobeying and getting away with it.

I'm not talking about handing people these Keys out of the rule-book. They should totally be made up when the person takes them. I'm talking about, when the guy says "My guy should have a problem with authority" saying "Okay, take a few minutes and write up what you, personally, think a problem with authority should be about. Then play to that."

Or we could just write down "Annoyed with Authority" and start framing scenes about that very subject (to record underneath that heading). Isn't that faster? And it does pretty much the same thing?

I mean, if you're like "I'd rather have a reward mechanic tied in," then fine. But there's different ways to do that. Give people XP for completing a bunch of scenes in which they rebel against authority. And then you don't have to spend the prep time. You just play and you get XP for doing cool stuff, whatever that cool stuff is, instead of doing specific kinds of cool stuff that you happened to plan for in advance.

I like this tag-team thing, actually. Tony's tough enough to take it and it gives us 3 different perspectives on this same experience we had.

I'm kind-of-sort-of doing something a little like this in Errantry (which is based on TSOY). As well as Keys, you have Obligations, which you get by being bloodied in Bringing Down the Pain. When you fulfil an Obligation you get XP. I haven't decided yet who chooses the obligation you get - the player or the Tenan (temporary GM). You have to fulfil all your obligations in order to consider your quest complete.

Hmm, if I tied them more tightly to Boons (which you must get a certain number of in order to complete the quest...).

Hmmm.

(They are mechanical, though.)

Ah, now I see our communication gap, which is that we're assuming a context you don't have. Let me describe with my example---

goal: Find New Teacher

• I'm Outta Here (got into a fight with his old teacher and sailed off to look for another)
• Caught in a Storm (in the midst of sailing, ran into a huge storm)
• Kayakwrecked (his kayak is flipped and tossed by the waves, he ends up lost at sea)
• Try Again (after being rescued by pirates, and working on some other arcs, he leaves the pirates to keep searching)

So, this is being triangulated by several other things.

First thing: I have four Path slots; this is assigned to my Water path, which tells me, approximately, "Positive experiences on this path are linked to dealing with other water Tribe characters, or behaving receptively/adaptively; negative experiences are associated with freezing up, isolationism, stubbornness." It also tells me, "Air Path scenes segue into scenes for this path, and they lead into Earth Path scenes."

This gives me two other points of reference - what's going on in my Air Path (Make A Friend) and my Earth Path (Become A Pirate). That tells me what to frame out of and toward.

Finally, I have the starting-point and stated ending-point for my Path, which tell me my initial state and my character's initial intent.

So we don't have just one lodestar here; there's a whole constellation of information telling me where to go. Because I've got all this data here, I don't have an enormous amount of wiggle room per scene, but each scene immediately colours the next scene; on the whole the system is extremely limber and responsive.

(Now, regarding what I said earlier re: constraint vs. weight; I think that Keys are weighty but not constrainty; I don't feel like they guide my play effectively because they don't really contain data.)

So that's where it wins over Keys for me; I don't ever put something down on my sheet and later say, "This isn't cool;" because of the way that the system tracks exactly what I'm up to and constrains me based exactly on that, it's impossible for something on my sheet that's currently relevant to be uncool.

edit jon why do you use html

I'll collate my path scenes together in a moment so you can see the way that they influence one another.

Tony,

First off: yes, I agree that constraint is a great tool. I think that to suggest otherwise would be silly.

That said, constraint isn't binary. You can have more or less of it, and just how much you want is going to depend upon your goals. For some purposes, the level of constraint imposed by Keys is perfect. In others it's too great (where I change my mind at the last instant about how I really want things to turn out), and in others it's too little (I'm imagining some situation in which having three options isn't focused enough).

On top of that, constraint doesn't have to be mechanical. One of the reasons the game in question worked so well (or so I posit), is that Jonathan, Shreyas, and I are all terribly Avatar fanboys. And we'd spent a fair number of hours on the drive down to Gen Con talking about the show and what it meant to each of us. We didn't need the mechanics to provide constraint, we had shared color and setting instead. For our group, for that game, for the specific goals in play that we had that shared understanding of Avatar provided precisely the right amount of constraint for us. Mechanical constraint would have been more constraint than we needed for our purpose, and thus would have reduced our flexibility to no advantage.

Thomas

Jonathon: I get the importance of the chaining. Or, at least, I think I do.

But it sounds to me like you're not comparing apples and apples here. You're talking about (on your side) a chain of freeform keys. But then you're talking about (on my side) an unstructured mish-mash of mechanical keys.

Do you think the same issues arise in a comparison between a chain of freeform keys and a chain of mechanically supported keys?

'cuz I certainly like the chaining. I just like the key mechanics too, and I'm wondering whether there's a reason I need to decide between either cake or ice cream, instead of just having my cake with ice cream on top.

Cool. But now you've gotten me all hungry, and I do not have either cake or ice cream in my actual, physical presence.

Oh, I totally need such tools to talk to myself.

If I sweep the dining room I get a chocolate covered pretzel. How are Keys any different?

I'll look harder for opportunities if a pretzel is attached to them. That's just how I am. Plus, I'm forgetful. So it's also kind of halfway a "to do" list. "Oh yeah! I thought it would be a good idea to secretly run out and disobey my master ... and here's a great chance!"

Keys are never a reward just from yourself, and they are never a reward just from someone else in the play group. Everyone always has the nuclear option, at least, of walking away from play if you use that boring, or offensive, or broken key. They are always a reward from the whole group.

But um, yeah, back to the topic. I guess I am not totally clear on how this chains thing is expressing something more generally useful than "have a goal, and drive toward it with your actions."

You know, reward mechanics are rewards only because of the social reinforcement of the other players. Suppose you're in tenth grade and you have this 27th level anti-paladin and he has 273 hit points, a -9 AC, and an unholy avenger that drains 1d4 Wisdom points from anyone you hit with it, and you just earned 45,000 XP for soloing Bahamut. Those numbers are a reward only because the other members of your gaming group are jazzed and energized by it. But if you tried to tell your mom about it, she would say "That's nice, dear", and those 45,000 XP are no reward at all.

So a reward mechanic doesn't need numbers -- what you really need is a way to establish standards of goodness, and a signal for the other players to communicate (dis)approval to each other. For an example I fully understand, in Court of the Empress, there's no numbers at all. Instead, the two reward cycles work like this: the courtier players are rewarded for learning how to please the Empress player, by getting to play more. And the Empress player gets to choose which players are most pleasing to her, making her social environment more pleasant.

To me, SJT's key chains seem like they ought to work really really well, because story-structure is an inherently pleasurable thing (it's one of the reasons why improv is possible at all). One thing that I want to ask you guys is whether only you can frame plot-arc scenes for your character, or whether a scene that someone else framed can advance your own arc. I'm guessing(?) it's the latter, because that would let you do cool things where in the same scene Buffy falls in love with the Boy, and the Boy acknowledges his demonic birthright, and all that braided plot stuff.

I should really talk to Selene about this to find out what you were doing.

Posted By: misubaI guess I am not totally clear on how this chains thing is expressing something more generally useful than "have a goal, and drive toward it with your actions."

This lets you have multiple goals (major, minor, etc.) that are being developed at different rates all at the same time, and yet are easily kept track of. And it supports instances where driving towards your goals leads somewhere completely different (unlike, say, Motivations in Exalted 2, which is an attempt to do exactly what you describe).

Neel: Yes. We didn't actually have any multi-PC scenes, but, in such cases, each character would certainly be working on there own chains, getting different things out of the same events. This, I think, could serve to drive inter-character tension and conflict.

Hi All. This is Brandon - as mentioned in John's post (way up there) and I would like to toss out some thoughts. Just a quick bit of background. I have played my TSOY character for 8 sessions now, with the goal of "Earn so much XP, the other players wonder why their characters aren't as cool as yours." I love XP. I love advances. I love telling my story. Narration is great and Story creation is fun, but so far, nothing has beaten the feeling of the climactic scene where I got to hit each and everyone of my keys and ultimately get to call out ... " and that earns me 22 XP." (out of a maximum 25.) And I all I had to do to earn that 22 XP was to take actions that lead to having my girlfriend be killed by my best friend, while renouncing my house (because I refused to kill my best friend. Long story.) Yeah, that's all.

I love the idea of the Key Chain and there is a lot of potential here. That being said, I do not see how your example fits the Key Chain steps you laid out.

>1. Declare and Record a Goal - GOAL: Find New Teacher
Sounds awesome.

>2. Frame a Scene Where You Make Progress Towards that Goal
>3. Record the Result

>SCENE 1: I'm Outta Here
Great. Begin the journey.

>SCENE 2: Caught in a Storm (Water Path)
>SCENE 3: Kayakwrecked (Water Path)

Here is where you lose me. I see how cool the storm is and that it is a part of your journey. However it doesn't really have anything to with your explicitly stated Goal of finding a new teacher. To my mind, the next scene relevant to your explicitly stated goal would be for him to meet the Pirate Captain (or whoever) and had a scene where he tries to decide if the pirate is the right person to become your new teacher,

My final point for now, is that your Element Matrix will create an awesome journey, but it does not really address the topic of achieving goals, which is what the TSOY Keys do. When I choose a TSOY Key, I am stating that my goal is to fulfill the 5 XP condition at some point. Sometimes that 5 XP condition is: Key of the Revolutionary (Strike a Major Blow against an oppressive organization) and sometimes it is: Key of Thrilling Heroics (Doing something thrillingly heroic, that could get you killed, while carrying the princess.)

Joshua: Yeah, basically. Writing down does several things. One, you can have mechanics built on that, so your system is powered by what you said is "just roleplay," which I think is pretty hot. Two, it really provides structure and pacing to play like you would not believe. When you can see what has happened and for what reason, it really helps you figure out what the next step should be. It makes the structure and purpose of play much more explicit. It's sorta like writing an Actual Play report as you go, with that sense of self-reflection.

Brandon: Yeah, I was actually just thinking the same thing. Shreyas was not really hitting his Chain theme as much as hitting his element in the early play we were doing. That's the problem with talking about this technique when we're only beginning to develop it, I fear. The suggestions you have about directly addressing his goals are a lot more on the ball, less filler.

Posted By: JonathanWalton

Joshua:Yeah, basically. Writing down does several things. ... it really provides structure and pacing to play like you would not believe.

Oh, I'll believe. FLFS gets tons of mileage out of the simple concept of talking about the game first.

Oh, I'll believe. FLFS gets tons of mileage out of the simple concept of talking about the game first.

And dude, when am I going to see this game? I'm getting impatient. Gimme gimme.

Posted By: JudaicDiablo
Here is where you lose me. I see how cool the storm is and that it is a part of your journey. However it doesn't really have anything to with your explicitly stated Goal of finding a new teacher. To my mind, the next scene relevant to your explicitly stated goal would be for him to meet the Pirate Captain (or whoever) and had a scene where he tries to decide if the pirate is the right person to become your new teacher.

I'd be a little worried about being too "on the nose" if we limit ourselves to only listing scenes that are directly, obviously and immediately relevant to the Goal.

When the PC decides whether the pirate captain is a good teacher or not, you must supply the audience (yourself and the other players) some reasons why he goes one way or the other. If we don't have any background context for this decision, we have to make it up on the spot, and that means doing some telling rather than showing. For example: "My PC considers the captain's offer, but decides that not to accept, because I'm pissed off at my old master for treating me like dirt, and pirates have to treat their victims like dirt in order to be pirates."

Now, suppose we have a storm and shipwreck, and the pirates pick up the PC. Next, there's another storm, and the pirates come through it unscathed because of the captain's seamanship -- establishing that he knows stuff the PC doesn't, and is a potential mentor. Next, there's a scene of the pirates robbing another ship -- establishing how they have to treat their victims. Now, when the captain makes his offer, the player doesn't have to explicitly say why his PC goes one way or the other, since the other players can deduce his reasons from the action. Pirates going through storms and robbing people doesn't look like it has anything to do with finding a new teacher, until the player MAKES it part of that.

What really jazzes me about the scene chain idea is that you can do these scenes and basically make a promise to the other players, "Trust me, guys, this will become relevant." And since the other players know what you're trying to do, they can help you by holding you to it.

Posted By: Matt Wilson
And dude, when am I going to see this game? I'm getting impatient. Gimme gimme.

November 16th release, with pre-orders and preview sneak-peak pdf in October. Which is all so freaking soon it's starting to wig me out.

Posted By: neelk

I'd be a little worried about being too "on the nose" if we limit ourselves to only listing scenes that are directly, obviously and immediately relevant to the Goal.

To Neelk: I do not disagree with you on this. In his specific example, the storm and kayakwreck are superfluous to his goal. I like your counter example though and those scene would contribute to his goal If we were to build a TSOY Key around it, it could look something like this (off the top of my head):

Key of the Searching Student
You are in search of a new master. You do not know who this person is or what he will teach you, but you will find him. When you take this key you may specify a person as your potential master for the purposes of this key. You may only have one potential master at a time.
1 XP: Have a Scene with your potential master (max 3 per session)
2 XP: Have a Scene where your potential master's actions sway your decision
5 XP: Risk great harm to win your potential master's approval
Buyoff: Accept a New Master or abandon your quest for a New Master.

In this case, I would record any scene where you choose a new potential master or earn the 2xp/5xp condition. Then things like sailing a storm and raiding a village matter.

Posted By: JonathanWalton

Um, but doesn't the whole categorization of scenes promote future scenes that are based on previous scenes as well as the categories themselves? That's like the core of the mechanic.

To Jonathan: Not necessarily. The entire basis of story creation is about building on what has previously happened. Categorizing scenes is, IMO, an unnecessary step. I can see it being useful in terms looking back over what you have done and saying, "aha - I want to build on that" but it is ultimately an aspect of storytelling that happens naturally. Players who take better notes often get to spring something fun from the games "history" but any could do that, if they remembered it.

That being said, I will now negate that statement with this. Take a look at Art, Grace, and Guts (http://artgraceguts.pbwiki.com/ - a Vincent Baker game in the works) and look at the "We owe" system. This seems like the type of codified example of what you are describing. My opinions about the system at this point are not firm, but if you haven't seen it, you should.

Oh, and I'd like to add that it was exactly that session and the way I burned through Keys to architect a massive revolution in the character that has me excited about this Key Chain idea. Can't wait to see it in context, Jonathan!

Mike: If you want to look at what a full-on game system based on this concept might look like, my hack of Exalted is beginning to come together here and here.