In Final Hour of a Storied Age (get the latest revision from the download page, or here's a link straight to the rev 0.67 PDF) I'm using freeform traits to guide narration, but I'm trying to use a trait/dice economy to prevent some of the problems I've encountered in other games, such as:

Trait spamming: Using the same trait in the same way, over and over again (often because it's mechanically the most effective)
Trait grubbing: Trying to include as many traits in the narration as possible to get a mechanical benefit, sometimes straining the limits of credibility and undermining the fictional integrity of the game

The mechanics are defined in Part IV: Playing Out a Chapter (and interact with Part III: Starting a New Chapter), but I'll try to give a brief overview: Two players are primarily involved in each chapter, the viewpoint player and the adversity player, and they go through a structured back-and-forth. Each one has story dice that they can spend to activate traits, which assigns an action die (the die size is based on how the trait is activated, ranging from d4 to d8) to the trait and puts it into a pool of traits that are available for narration. If a trait from the active pool is brought into the narration (a la DITV) the player picks up the die to roll for an exchange. The player with the highest individual rolled die wins the exchange and can bump up the die size of one of the traits he rolled (to a max of d8). The loser counts up how many of his dice are lower than the winner's lowest die, and must deactivate that many traits from the traits that he rolled from his active pool. Aside from the winner's die size change and the loser exhausting some traits, the traits stay unchanged in the active pool. (This continues until one player runs out of story and action dice, or until one side wins three exchanges in a row.).

The mechanical effects I'm aiming for here:

The risk of getting traits exhausted, and the different risk/reward profiles of rolling single vs. multiple dice, are intended to counteract the trait-grubbing that can happen if there's a linear increase in effectiveness with each additional trait brought in. I'm hoping that there are reasonable mechanical arguments for using one or two (or maybe three) traits on a single roll. I want the mechanical incentive of choosing one option over the other to be weak enough that you don't feel stupid for choosing a “suboptimal” one because you have a cool idea for narration that uses a particular combo of traits, but strong enough to give you some guidance for what to narrate if you don't have an idea.

Setting the die sizes of the traits on a per-chapter basis, the loser-exhausts-traits mechanic, and the fact that you don't get to bump up a die size when you win with only d8s (d8 is the max, so you can only bump up a die size if you include a d4 or d6 action die in the roll) are intended to encourage variety in the traits used instead of spamming the same ones over and over again.

I'm looking for feedback on whether my mechanics seem like they're doing what I want them to do. For example, is there a powerful strategy in the dice game that I'm overlooking that would make the trait choice in the narration feel highly constrained? In the recent playtests (check my blog or podcast for more detail) it seems like the players are favoring single large-die rolls, but I'm not sure if the system is pushing them there very strongly, or if that's just one of multiple valid approaches. (Obviously these mechanics are tied into the rest of the game, so a broader discussion might be necessary, too. Also, I'm interested in ironing out any unclear or confusing aspects of my rules text, so if there are things in there that are hard to follow ask questions or point them out.)

Hi Dan! I've been looking forward to this.

Right now, my concerns are awfully basic, almost kindergarten.

Do I understand correctly that a character can draw more than one trait into the conflict? I'm 99% sure that's the case. If so, then I suggest that one solution may be simply to limit how many. Many people completely miss the fact that in The Pool, for instance, you choose one Trait to contribute dice to a given conflict, if any. The others on the sheet may be included in the narration of play, but they will only be Color.

There is a design feature that has persisted in the indie RPG community for about five years in an uncritical way, almost as a fetish rather than a genuinely useful technique. It is: "say something, invoke a trait, grab a die (or bonus, or whatever)" as a subroutine to conflict resolution. It works very well in several specific games: Hero Wars (HeroQuest), Primetime Adventures, Dogs in the Vineyard, and if I'm remembering the system correctly, Polaris. It is horrible ass in many others, mainly due to the hawtness of a truly wretched game design called Wushu.

What I'm saying is that you don't have to have all these traits leapin' into the conflict in mechanical terms if you don't want. Right now, it seems as if you're offering an incentive to do it (the mechanical advantage), then, finding that you don't really like having them all in there, trying to offer a disincentive for doing it too much. Which means you're designing against yourself, creating what may well be a hunchback. Sure, it'll have "story game" bragging rights because you talk! you get dice! you talk! you get dice! et cetera, but it'll suck. Why? Because you're throwing people into a competing incentive-disincentive mind-set, which only ever results in trying to cheat one's way out of it.

Markus began a series of very intense threads about traits that may be too much to dive into,* but at one point in there, I broke down a bunch of different systems in terms of what resources trait use depended on. Sometimes it was a how-many thing as in The Pool and to a lesser extent PTA; sometimes it was use'em-up thing as in Legends of Alyria, 3:16, and Polaris; sometimes it was an ordering thing as in The Exchange. A lot of people found that they were playing games without using the resource limits for traits, explaining why they kept running into exactly the problems you listed (myself included, concerning Hero Wars).

I bring up resources specifically because they solve the hunchback problem - on one side, you have a mechanical incentive to use the trait; on the other, you can only use so many of them or so often. There's no competing incentive problem. That's why I suggest that instead of trying to create a balancing act between incentive and disincentive, which I frankly think is impossible, you consider a resource limit instead.

Best, Ron

* Can someone explain the true reason behind "traits" (PtA style) to me?
[Space Rat] Femme babe action at GenCon
Traits and the darkness that comes before
[Legends of Alyria] Traits! Traits!
(update: I no longer think my proposed before/after distinction in those threads is correct in terms of either/or game design; I now think that's a consequential Ephemeral phenomenon that bears more analysis)

I haven't read the linked threads yet, so I may have more after I do (holiday preparations might mean it will take me a while to get to them).

First, to clarify the way my game works mechanically: The current rule is "use as many of your active traits as you want in your narration", but there are different pros and cons to using one, two, three, etc. traits, and there's not a lot of mechanical reason to use lots of traits. You need to use at least one, or else you have no dice to roll. The way traits are used is similar to raises in a DITV conflict (I tend to avoid using the word "conflict" when describing my system since there's no "freeform roleplay for a while until we realize there's a conflict" step, and the word conflict sometimes makes people assume that's there), but the traits don't disappear from the pool when invoked, and you need to invoke at least one trait for each exchange. These traits can come from a single character, multiple characters, or environmental threats (it's most common for the viewpoint player to only use traits from his own PC and for the adversity player to use a wider variety of stuff, but the viewpoint player might also bring in traits from travelling companions and the adversity player might choose to use a single PC or NPC as the source of adversity in a chapter). Traits are sort of resource limited since it costs you a resource (story dice) to put them into your active pool, but not in a completely predictable way since you only lose them out of your pool based on certain die roll results.

I don't think that a hard requirement on the number of traits per narration chunk feels right to me. I think that will make the narration feel less organic. Especially if the limit is one trait, I think the step "narrate the adversity" or "narrate your response to the adversity" would become functionally equivalent to "pick a trait", which would make it easier to slip into "weak trait invocation" territory, where the narration becomes a mere formality and players would just say their trait with an implied "it's obvious how that applies". I think that the option of having a variable number of traits based on your narration makes it more likely that you'll mentally categorize this step as narrating something rather than a purely mechanical choice. (This thought isn't fully baked yet -- I haven't really tried to articulate this before, so it's a good question). Having the ability to invoke multiple traits in a single narration also means that players don't have to worry about trait overlap, breadth and depth, etc. Having both "Master Swordsman" and "My Family's Heirloom Blade" traits can be interesting on the same character because they would shade narration in different ways when used independently, but if you couldn't use them both at the same time you might be discouraged from taking them both as traits. There might be more reasons that I want to allow multiple traits per role (it's a design decision I made intuitively, not deductively, so I'll probably never be sure what all of my reasons were).

I'm not sure I understand the "designing against yourself" point. From my POV, what I'm trying to do is make different mechanical choices (which correspond to different narrational choices) seem valid. If the mechanics push overwhelmingly in one direction, you either get people only going in that direction or else feeling stupid or guilty when they don't.

When I first read your post, Ron, I started thinking about how to articulate why I'm using freeform traits in the first place, but on rereading I don't see you questioning that as much (maybe some of that stuff is in the linked threads?). Let me know if you'd like to talk about that.

First hand experience with playing Storied Age is that the advantages and disadvantages of using few vs. many traits are balanced enough that I'm comfortable using whatever combination of traits works best for the narration I want. This might because I'm missing some exploitable feature, but nothing about the game makes me inclined to look for one.

Hi,

I want the mechanical incentive of choosing one option over the other to be weak enough that you don't feel stupid for choosing a “suboptimal” one because you have a cool idea for narration that uses a particular combo of traits

I think trying to make all options roughly the same but not quite, still means your making sub optimal choices if you look at it from a pure mechanics/hardcore standpoint, it'll just be less of a sub optimal choice.

(d) How do you choose which trait to use, among the list of those your PC possess?
  - I'm trying to obtain maximum (mechanical) effectiveness
  - I'm trying try to reveal something new about my PC
  - I'm trying to bring to the table the theme/issue that I hardwired to my PC by choosing this trait
  - I'm trying to choose the trait that preserves the most stringent consequentiality/coherence of the fiction

Dan,
  I just finished reading this.
  The writing is definitely punched up. It has a much more friendly tone and it is just easier to read.
  Here is how I read the new mechanics:
Roll Lots of dice:
Harder to narrate in so many traits
Easier to introduce more characters
More story dice to activate
Exhaustion will be harder to recover from
Increases chance of rolling highest
Increases chance of winning a tie
Increases number of dice that will be lower than the oppositions dice if you lose
Increases chance of rolling a die that is too low to exhaust the oppositions traits

Roll Less dice:
Easier to narrate in traits
Harder to introduce more characters
Less story dice to activate
Exhaustion will be easier to recover from
Less chance of rolling highest
Less chance of winning a tie
Less dice that will be lower than the oppositions dice if you lose
Less chance of rolling a die that is too low to exhaust the oppositions traits

None of these are guarantees though, so it feels pretty balanced. I think because of the increased cost in player effort and resources, I would probably go with less Traits, but that is just because I like to play cautiously.

I finished reading the last of Ron's linked threads. I think the before/after paradigm that Ron mentioned having evolving views on isn't a great model. I designed my game after listening to and reading a lot of Vincent's clouds and boxes stuff, and my traits are intended to work as prompts for narration by being pattern-matching operators pointed at the narration. The traits themselves are sort of backward-looking in that they only kick in once someone has narrated something that applies, but they aren't as problematic as "yeah, I guess this trait probably applies to the situation we've been describing" traits because the person providing the narration knows before they start speaking that the narration they provide will need to hit one or more of the traits in question -- when the sender knows what kind of patterns the receiver is configured to receive it's easier to send a clear message. So if you and I both know that I have the traits Master Swordsman and Booming Voice, it will be relatively easy for me to give narration that will satisfy one, the other, or both of those, and it will usually be pretty clear in context which traits I'm trying to invoke. If it isn't clear, the receiver isn't put in the confrontational position of saying "you can't use your trait", he's able to employ the more socially harmonious "I don't really get how your narration hits that trait, could you revise the narration to make it clearer for me?". Since the narration is obviously bounded and still tentative (basically, just what you've said "this round", like a single raise or see in Dogs) it's easy to ask for the narration to be edited, whereas it can be much harder to ask for the editing in a more diffuse "whenever it applies" trait since that could potentially impact large swaths of seemingly established fiction.

I think it's also useful that the narration I'm asking players to do isn't solely focused on invoking the trait, but is also targeted at an orthogonal pattern-matcher: either providing adversity to the viewpoint character or responding to that adversity. By needing to serve multiple masters I think it's less likely that players will perform the functional equivalent of just reading off their trait -- you can't just have your character shout to use Booming Voice against me, you have to narrate something that both demonstrates the Booming Voice and is an obvious source of adversity for me, like shouting to drown out what my character is trying to say. This is pretty similar to things like Mouse Guard's "describe how you're using one of your skills to give a helping die", but since the actual narration is more important (since it guides the viewpoint player's response to the adversity and the narration of the results of the exchange) there's less likelihood of people just saying "I'm using X" without actually giving the narration that describes how they're using X.