Monthly Archives: July 2016

2d6 and 3 out of 5

On a very primal level, one of my favorite things about the various Powered by the Apocalypse games is that they use 2d6 for resolution. It’s just about the simplest possible way to generate a curve, using the most ubiquitously available dice. The math is easy and fast. It uses an even number of dice, which may seem like a small thing, but is important when you want to buy cool dice (you don’t want to know how much I paid for the metal set I picked up at Gencon).

But PBTA was far from the first 2d6 system, and and I can’t think of another such system that has been so sticky. A big part of that is, I think, that PBTA has driven forth a very concrete means of reading the dice that only generates 3 outcomes1 with very little player training. Once the player has internalized that 7-9 is the middle result, they can easily infer whether they’be done better or worse.

There’s some interesting knock-on effects to this. First, it rewards the simplicity of the dice. It would be easy to produce a more nuanced curve with more or different dice, but that could impede the ease of learning that core number. It would also introduce a temptation to get fiddle, which would also detract from the simplicity.

It also allows for the designer to play GM in a very concrete way. The design of a move is a decision bout how an activity should look – that’s one of its big strengths. It’s apparently simple design actually conceals something a little more complicated. I’ve talked about this before, but the most direct way this is expressed is the range of outcomes. Consider:

10+ Success with Benefit
7-9 Success
6- Complication


10+ Success
7-9 Success with Complication
6- Problem


10+ Success with Complication
7-9 Not Quite Failure
6- Ha ha ha ha ha

As you look through PBTA material, you can find all three of these patterns (and others) reflecting the designer’s take on that particular sort of action in the context of the genre. Sometimes this is applied with grace and nuance, sometimes it’s shotgunned across things as an expression of the designer’s taste. Which is fine – PBTA is an opinionated game system, so this is a feature.

But, of course, it makes me go hmm.

If one wanted to start from the core idea (a simply expressed narrow range of outcomes) it would probably be possible to create an ur-list of maybe 5 outcomes. You could theoretically produce three sets of moves out of it (high, medium, low) out of that set. So, suppose the list is:

  • movespectrumSuccess with Added benefit
  • Success
  • Success at a cost
  • Mitigated Failure/Complications
  • Disaster

So a “nice” move would be

10+ Success with Added benefit
7-9 Success
6- Success at a cost


And an average move would be

10+ Success
7-9 Success at a cost
6- Mitigated Failure/Complications

while an “unkind” move would be

10+ Success at a cost
7-9 Mitigated Failure/Complications
6- Disaster


Such an approach would not need to replace any existing moves so much as simply provide a set of guidelines for whipping up moves on the fly. And that works on paper, but it has one big drawback – it is effectively a back donor for inserting difficulty into PBTA and that is probably a very bad thing indeed.

So I’m not so sure there’s much use for this idea in the context of PBTA as anything but a shiny bauble. However, as an indicator for how to maybe steal a chunk of PBTA tech and use it elsewhere, this may prove a very interesting starting point.

  1. Yes, there are ultra-positive results in some builds, but they are something of a sidebar. ↩︎

Signifier Stats

A lot of games have a category of stats (often called something else) which may have some mechanical effect, but are most critical as signifiers of player intent. I’m calling them signifier stats because I need to call them something, because I’ve found myself thinking about them a bit.

The most famous and obvious example of a signifier stat is alignment in D&D. It’s not a perfect communicator, of course, but it does a decent job of conveying the kind of play that the player is looking forward to. Choosing Lawful Good is not (usually) an indicator of an interesting a shady political play or bloodthirsty mayhem. It’s incomplete – sometimes an alignment is chosen for contrast (the one honest man among thieves, or the secret betrayer) or because the player has seized upon a particular nuance (The paladin who is super lawful, but who does not recognize the current ruling power as legitimate, or any number of doomed heroes). But in any case, alignment is n obvious part of the communication, and when your whole playgroup chooses Chaotic Good, they’re telling you something.

A less famous but even more clear cut example is Good and Bad Stuff from the Amber Diceless RPG (and its successor, Lords of Gossamer and Shadow). Amber characters are bought with 100 points. Unspent points translate into “good stuff”, and you can overspend a little an pay it off by taking a few points of “bad stuff” . Effectively, if you spend 105 points on stats and powers, you have 5 points of bad stuff. If you spend 95 points, you have 5 good stuff.

Because Amber is a diceless system, “stuff” is where a lot of things dictated by luck come into play. Lots of good stuff? Things go your way. Bad stuff? It can feel like the universe has painted a target on your back. This was not entirely reliable – for many GM’s good stuff merely meant terrible things only happened frequently rather than constantly – but it established a spread of expectations at the table.

When the game first came out, I thought of this as a balance mechanic. It was tied into points, wasn’t it? Clearly, the point was to penalize players who overspend and reward those who didn’t. But in time I cam to realize it was and of the reverse – players taking bad stuff were agreeing to make the GM’s life easier and were getting rewarded for it. Players buying Good Stuff were buying a degree of protection, and paying for that1.

Now, collecting player intent before play starts is one of those things that it’s just worth doing. There’s no substitute for talking with your players about expectations before starting play. But unless you are playing a very thematically constrained game, that conversation will probably reveal a spectrum of possibilities that do not necessarily indicate where each player wants to go. Even if everyone agrees they want swashbuckling high adventure, having an explicit indicator of the kind of play each player is going for within that space can be a super useful idea.

Obviously, this purpose can also be served by direct hooks on the sheet – things like backgrounds or aspects – but those may not necessarily answer the same questions. Signifier stats occupy a space between the tone of the game as a whole and the specific character hooks, and used properly, they offer a bridge between the two.

To illustrate using an amber example, let’s say we have agreed on a nasty, political-and-fisticuffs game of Amber. I’ve made my character and bought a connection to my parent, one of the princes or princesses of Amber.

If I am playing a Good Stuff character, my relationship with my parent is probably pretty healthy. They’re quite possibly still a jerk, and they may yank me around of army own good, but I know they have my back. If I buy more allies over time, they’ll be reasonably solid – betrayal is a risk, but it’s the exception.

If I’m playing a Bad Stuff character, my parent may very well consider me one more playing piece. I benefit some because I’m a useful pawn, but we both know this is a relationship of expedience. If I gain allies, I must constantly watch my back, since I am in a nest of vipers. The occasional person I feel I can trust is a treasure (and a liability)

I could explicitly articulate these things with each connection and with each relationship, but with a signifier stat, I don’t need to. That saves hassle, yes, but it also means that players who are uncomfortable explicitly seizing the narrative can still express a preference. That’s some pretty useful mojo.

  1. But, critically, the game never said this was the case. Many players picked their Good and Bad stuff rating for entirely non-mechanical reasons because of the type of character the played ↩︎

The Size of Aspects

I’ve remarked on a few occasions that an essential decision in Fate is how big aspects are. That is, should aspects be the big, important signifiers of what really matters, or should they be more of a language to express the moment in mechanical terms.

As Fate has evolved, it has gone more down the second path, and there are a lot of benefits to that. It is fairly easy to turn any scene into a handful of aspects, and by extension have the baseline mechanical attributes you need to run the scene. That’s a useful tool.

But it’s important to note that plenty of other system can do it. If I’m doing a Risus or WaRP variant, I can assign d6 pools to any arbitrary thing you can think of (On Fire 3d, Deep Shadows 2d). In Cortex Plus I can do similar things with a die value (on Fire d8, Deep Shadows d6). If the game has reasonably flexible descriptive mechanics, it’s pretty doable.

In this context, the main feature of Fate is that it’s simple. In each of the examples above, a little bit of thought needs to be given to assigning a rating. Is this a 3d6 fire, or a 4d fire? What’s the difference? Sometimes that can be easy to intuit (especially if you know the system very well) but it’s still an extra step. In Fate, it is simple On Fire or it’s not.

Obviously that’s double edged. There are times when you might want granularity, but that is the nature of tradeoffs. The right answer is the one you need right this minute.