I know I started something yesterday and it’s frustrating to interrupt it, but I’m reading Herman Melville at the moment, so I’m in the mood for ten thousand word parentheticals.

I got an email the other day and I wanted to react to it in depth and publicly because it probably voices the sentiment of a lot of Diaspora players and so the default audience for anything new that VSCA will produce. I won’t reproduce the email (it was sent as private communication after all) but here’s the gist: Hollowpoint seems like a cool place to play (modern action) but the system is alien and not to my taste — please please do it differently.

From the general, abstract place in my head: Hollowpoint is an experiment and experiments need the freedom to fail. One of the things it experiments with is a kind of objective that is common in action scenes and badly modeled (sometimes impossible to model) in what I will call “guy versus guy” systems. So I’m going to try something very different (though not unrecognizable: leaf through your copy of Reign) to get at what I want to get it. I am certain that this divergence will be unappealing to a lot of people. That’s cool — that’s data. It’s also really appealing to at least one person so I hope there will be others. If you’re on the fence (and as the game does not yet exist, that might be a good place to be), hear me out. If you’re committed to disliking the very idea, move on — there will be other VSCA games and if you love Diaspora, you can already get that.

So I was watching Heat the other night — a Michael Mann movie with some very smart action scenes — and noticed how well Hollowpoint maps into it, and that’s exciting, because that film is very much in the target zone for the game. By way of example is the famous bank robbery scene: the crew has executed a bank robbery without violence and in the course of exiting they are bounced by the police. The crew has automatic weapons, great training, and willingness to cause harm and hurt but they are also professionals: their objective is to escape with the money.

No in guy vs. guy gaming, this is really, really hard most of the time. Because the system will focus on which cop your character is trying to kill each time-slice, you the player are focused on the wrong thing with distinctly uncomfortable (to me, and in this genre) effects.

First, I (the player) have to plan how to most effectively kill police officers because what the system primarily lets me do with my assault rifle is kill people. I am not enjoying that in this context.

Second I (the character) am not explicitly interested in killing police officers. I am interested in escaping with the money and don’t care if I kill police officers. But the system models me defeating police officers with my rifle.

Finally I (both player and character) have sophisticated, staged objectives that involve violence against a large opposing force with full knowledge that I cannot just kill all of them (and here’s a place where some guy vs. guy games really drop the ball for me — I can kill all of them. Seriously, I can kill the entire LAPD to solve a problem, just by looting corpses for ammunition.)

The scenario is a classic “breakout”. The police are technically a defensive surrounding force and the robbers objective is to create a weak point in their line, penetrate it, defend their egress, and escape. People are going to get killed, but the solution is not about killing people. You don’t create a weak point in a defensive line by killing everyone — you create it by making a zone where no defender is willing to oppose you effectively. If they are all dead, that’s certainly one solution, but you, with the objective of breaking, don’t actually care. And if you’re a pro you also know it’s not a feasible step in your plan anyway.

A breakout is achieved by aggression. The unit under siege identifies a point of egress and advances on it, concentrating fire. Flanks are protected to avoid being enveloped but the focus of fire is the point of egress. And you advance constantly and aggressively. Go watch Heat and come back.

Okay see that? That’s what you want. And when the line folds, you exit, secure transportation, and depart. The criminals are using several important tools in this process: they are making people feel too afraid to be effective by shooting the shit out of them. Terror is the tool there. They are identifying and neutralizing core sources of resistance (vehicles, commanders). Killing is the tool there. They are leveraging the fact that they do not care about innocent bystanders and the police do, giving them vastly more free mobility and fields of fire. Again, this is mostly about Terror.

But the bulk of it is not about a series of guy vs. guy incidents. It’s about effective use of ammunition, mobility, aggression, planning, knowledge of the space, sustaining fire (rapid reload!), and effective fire (shooting at the target — a notoriously hard thing for non-sociopaths to do). So a system that gives you a tool for defeating one other person by intimidating or killing her is not giving you enough to work with. The richness of this scene — and all of its energy — would be missed by focusing on who shot who. Watch that scene again and listen to it. This is one of a very few films that use accurate sounds of gunfire. Turn the volume up. Listen to the difference between the light assault rifles of the crew and the boom of Pacino’s heavier rifle. Listen to the echoes off the buildings. The chief issue resolving this scene is how afraid everyone and how willing they are to do harm. The ability to accurately hit a target is a tertiary factor at best.

So Hollowpoint, being interested in this sort of scene, does not do guy vs. guy action except as an exception. Instead it’s about the individuals in the crew and their contribution to an action against an opposing force with a common objective. An assassination, for example, is not “killing a guy”. An assassination is a sophisticated preparation of a space in which an effective killing blow can be struck while allowing the assassin to escape. An ambush is not “killing six guys”. It’s again a preparation of space in order to destroy a unit of men (as a unit, not each man) and then exit the location safely (or otherwise manage the objective: you ambushed them for a reason).

Now I am not slagging guy vs. guy gaming. Diaspora is very much a guy vs. guy design and I love it. But the model doesn’t do everything well and it doesn’t do what I want here. So far, in play, Hollowpoint certainly does meet my needs. I know every roll that Val Kilmer’s character made in every scene. I know what choices he made with the dice he got.

Interestingly, the most disappointing part of that movie for me is the last half hour or so. I think it’s obvious why, in light of this discussion. It forgets what it’s really about. Or it doesn’t detect what I think it’s about.

It’s cool to dislike some or all of my games. Vive la différence.

–BMurray

The FATE system is a highly successful one in a relatively wide range of genres (basically anything where the characters are expected to be very competent — not necessarily superheroes or even heroes, but competent). It seems unwise to dick with something that works, but part of its success is its malleability — it’s not just resistant to dickery, but rather it thrives on it. When you get your teeth into it the temptation to bend it to your will is substantial.

Partly you bend it to your will in play. That is, it can be played extremely well by using the rules as written to produce unexpected effects. Weak players will succeed less often than strong players. A group with one strong player can dominate, tactically, a group with no such person. It’s a tactical system and it rewards mastery.

But that’s not what this is about because when we play we’re usually also designing and so we are bending it to a different will: the will to make a new game that uses (in the dirtiest senses) FATE to accomplish our own ends. I think that one of the deepest savageries performed on the system was our own game, Diaspora, in which our objective was to play the desperate, broke, beleaguered competence heroes of the Traveller universe using a game intended to deliver pulp superheroes.

Now with our new games we’re pushing even harder — hard enough that it’s probably not reasonable to call them FATE games. Certainly they owe some debt to FATE (some more than others) and certainly at least one will use FUDGE dice, but they aren’t really reasonably FATE games any more. But let’s look at FATE because the game that uses the most of FATE, Soft Horizon, is on my radar at the moment.

Much as we are often loathe to admit it, role-playing game systems are basically fluff surrounding a resolution system. You invariably make characters who are described by the elements that interact with the mechanisms of the game to say yes or no to, “Can I do …” questions. One or more mechanisms exist to answer the question. They bring character description metrics to bear on a randomizer (or not — the randomizer is part of the gears not part of the definition) to find an answer. Here’s FATE’s central machine:

My character is this awesome (SKILL) + this consistent (FUDGE dice) + because of these things (ASPECTS).

Yes I am ignoring Stunts. Stunts are interesting because they aren’t here. I may talk about them later. I may not. I have a love-hate relationship with Stunts. They are a kludge that works really well, and so I despise them aesthetically but I need them. I am convinced there is a better way but currently they are inadequately defined and so it’s hard to get a grip on improving them. I think I have a path. But later. Patience.

So the SKILL rank is how awesome you are. As a designer ¹ this informs us about the skill structure. Armed with this knowledge, we can make sense of skill pyramids versus skill columns versus unstructured point buy versus whatever other variation is out there.

The skill pyramid, for example, is several declarations. In the 5-cap we use in Diaspora (and I will maintain it is optimal and am talking here about why) it represents 15 skills which is almost half of the total set of skills. So, the bottom five are things you can do. You know about them. They aren’t your focus. Because there is a disparity of four points between top and bottom which is a BIG DEAL in FATE, you are going to try to avoid using them. This creates an interesting dynamic at the table — with Computer 1 I am not awesome at computers. Consequently, even though I can make sense of them, I am going to find a way to solve problems that doesn’t involve computers. If I have Stealth 5, I am going to break in myself rather than hack. That’s a no-brainer. So we discover that the lower ranks of skills are not so much opportunities for different kinds of action but rather reinforcement of the awesomeness of the pinnacle skills.

The mid-range is where a lot of incidental action occurs where we are not super interested in success. This is where we go when we are partially invested in success and think failure might even be cool. Or sometimes it’s where we have to go when we can’t find a story in the apex. Ranks two and three are interesting spaces. They often represent desperate, fish-out-of-water scenes. These are the places where the character has to act even though she is not competent.

The peak is where the play mostly resides. This is where a good player is always going to steer the action. A choice of Stealth as an apex for the pyramid is shouting at the referee, “I am a spy! I sneak! I will solve un-sneaky problems with sneaking whenever I can! Make me sneak.” It is a character class. These three skills (the rank 4 and rank 5 ones) are the core competencies of the character and therefore the key definition of the character. ²

These things are not core features of FATE. They are core features of the pyramid. If you don’t want this to happen, then you choose a different shape. If you do want these things to happen, this is the shape. So that’s SKILL. That’s what the character is good at and what she is going to use as tools when in trouble.

Now the dice. The choice of FUDGE dice, a range from -4 to +4 with a strong peak at the zero, is a declaration of consistency: characters usually get the results they expect. Sometimes there are crazy outliers, but for the most part you are going to get a level of success roughly commensurate with the skill you choose. Those of you out there who are still awake will note how this feeds into the behavioural stuff I just talked about with respect to skills.

So here’s another place to dick around that has profound effects: by changing the probability curve you change the reasonable expectations of the player and you therefore change the tactical value of skill selection. With a wider range of randomness it will be less relevant which skill you choose (if you have a range of +/- 100, to be absurd, it doesn’t matter what skill you use at all really unless there’s a crazy powerful peak at zero). With a narrower range (or no dice at all — try that some time if you’re the exploratory type) skill selection is critical. The FUDGE range creates some very desirable effects that are thrown out of whack with even minor changes. We experimented with all kinds of d6 variations but ultimately none of them rewarded good skill selection as well as the FUDGE dice while retaining enough uncertainty to be exciting ³. So FATE system gives us a certain quality of consistency with the FUDGE dice that has certain effects. In some games they are desirable. In some games not. This is a place to tweak carefully.

Finally, the most interesting narrative component of the resolution, which answers the question, “How come I am so awesome?” Aspects are tiny little stories about the character that the player can leverage for advantage (at a price) during conflict. Varying how many and how much they cost is a knob you can turn pretty eagerly without breaking the game and the effects are not all that powerful. You can yank this knob (LOL) all over the place and maintain a high quality game without even affecting the capacity of the system to deliver its intent. This is the least interesting place to dick around, in my opinion. But there are undesirable behaviours that you can add mechanism to avoid.

First, it kind of sucks when a player leans on  all of her aspects at once because she has the fate points to do it and the narrative is strained, but she did do the tactical work to flex the economy so that she had the power to leverage. It’s a tough balance, honestly, and it’s one that we chose to handle mechanically rather than leave in the hands of the referee: we introduced the concept of “scope”. Basically, any Aspect has a scope — a context in which it exists. So characters have ten Aspects and they are all in the same scope: the character. A scene might have an Aspect. An enemy might have one. A zone might have one. Each of these are scopes. Now you can say things like, “you may only tag one Aspect from each scope.” Now the player must search the environment for advantage. This in turn rewards the group for placing aspects on the environment. This rewards teamwork with success. In some games this is highly desirable.

In some games it is not. This is really important. Diaspora is not all games. It’s not the only game anyone should ever play. It’s a game in which it is rarely the case that each character has the opportunity to deliver a “killing blow” ⁴ absent the assistance of others. This undermines the “I hit with my sword” litany that some combat systems devolve into because the chance of success is actually very low if everyone acts alone hitting with their sword. Instead, characters play supporting roles, planting aspects in new scopes, so that one person can grab the spotlight (and not just the “I am narrating” spotlight that we usually mean, but rather the “I am the center of attention, the source of awesome, the leading role” spotlight) and deliver an awesome success to resolve the conflict.

That’s not always the story you want to tell.

So the way you handle aspects changes the tone of the game substantially.

Okay so there is a lot more to FATE, right? I mean there must be. Most FATE games run hundreds of pages. Well here’s the thing — we have no idea what FATE actually is. Right now, FATE is a series of divergent examples. So up there is what FATE is to me. It doesn’t need anything else to be FATE. How awesome + how consistent + how come so awesome. Now, it’s fair to say that practically every game everywhere uses some clear variation on the first two thirds of that. So aspects are what make FATE novel. That’s it. There are fiddly coolnesses out there like the Time Track and stress and consequences and all that, but they are mostly derivative ideas (derivative of aspects I mean) and decorations on the core. You want a FATE v3 SRD? Skill + Dice + Aspects.

Now I know I’m leaving out the fate point economy. A year ago I wouldn’t have but since we started doing the dice for AP promo I’ve read a lot of actual play and here’s a dirty secret that doesn’t get spoken a lot: the fate point economy is far more dependent on style of play than on mechanism. The same rules generate vastly different fate point behaviour at different tables. That suggests that it isn’t as important as we sometimes think, because it seems to be subservient to play rather than driving play. That’s cool — even awesome — but not what we thought was true at one time. And so I’m not going to draw out compels and fate point bidding and refresh and all that. It’s important, but it doesn’t drive play as much as we hope. It gets reliable results at a table you know, but at a new table under a new referee, different things happen. With this five guys, ten fate points never get used up. With this other five guys using identical rules, players are begging compels by the second hour. With another five guys (who shall not be named) there’s a subtle player vs. player action happening with players trying to manipulate each other using fate points as a social currency. The fate point economy is important, even essential, but its effects have too much variation to talk about with any authority. This is the knob you hand to the consumer of the game: twist this wherever suits you.

Okay, so Soft Horizon. Soft Horizon is the only game in the upcoming list that is certainly a FATE derivative, but I don’t think it’s FATE in the sense expected by the OGL for FATE. It will still use the OGL, but I don’t think we owe it to FATE. It’s still Skill + FUDGE dice + Aspects, though, so it’s that much FATE. Essential FATE.

More later: this is plenty long.

–BMurray

Getting six people to work on a focused goal is pretty easy — I’ve run teams plenty of times and it’s just not all that complicated. It works easily, however, because the social structure lets it basically be me with five extra sets of arms. That is, I plan and organize and command and review and collate and present. My team members get their piece done according to my plan. This is not collaboration. It’s a way that some things work, and it’s good, but it’s not collaboration.

VSCA projects, by contrast, are strictly collaborative. Each author has (theoretically) equal input and consequently organization needs to coalesce rather than emerge whole. This means that we need a richly iterative model in order to get work done. This was brought home for me recently while discussing Chimaera progress with JB, during which I tried to explain how we were going to bootstrap the idea into a game. All too often I have no idea what I am doing until I try to explain it to someone else, at which time I discover that I really do have a methodology.

The essence of VSCA collaboration is play. We never go away and write a game whole and then playtest it, but rather we playtest very coarse drafts and fragments of subsystems and revise based on the information from that and over and over and over. Always playing. We don’t have meetings, we have game nights. But we’ve had precious few game nights in the past two years that haven’t also been game designing. I don’t think I can go back to “just play”. This is play for me now.

So we don’t really do anything with a project until someone writes enough to play with. It can be just a subsystem or a dice gimmick, but it has to be enough to drive an evening’s play and discussion. So we start with this kernel of a few thousand words in which someone tries to explain the game. It doesn’t need to be refined and it doesn’t need to have a “voice” yet, but it does need to convey enough to everyone else that we can all get our teeth into it. Until this happens there is no project. I don’t care how much information you have in your head, until you share we aren’t collaborating.

Once we have that kernel it goes up at the wiki where anyone can hack at it. Now all kinds of semi-organized work happens. Links to AP reports and audio go up. New text goes in. Micro-fiction from play or imagined play goes in. Proposals are made. Things get re-organized. Images and other art tests get made and linked up. Basically the wiki becomes a multi-media creative collage of effort from the collaborators. It’s a mess. Probably no one can play with it but us. But we certainly can.

Every game night that is exploring a project is recorded. Immediately after play, someone tries to capture the essence of what we learned that night and post it up at the wiki. Early on there are more questions than rules, but that’s cool. We are confident that an organized project will show itself over time. We don’t need to make it happen yet.

The basis of iterative development is obvious — the idea is that you cannot design perfectly and then implement. That is, designs are always flawed. With that as your core premise, one solution is to design only a minimum and then get it working in as real an environment as possible, take notes, and push those back into design. Revise the implementation and re-test. This avoids designing shit you don’t use and highlights stuff you didn’t know you needed. Anyway, all of this is familiar to software designers and fits nicely into the led-team model.

It also works for peer collaboration. In fact it works better for an important reason: it establishes a certain amount of investment (even ownership) in all the collaborators. The core idea will change — although that came from one person’s head and was perhaps perfectly clear in there, in play and revision the author’s peers will take what they like and expand it. They will tear down what they don’t like. They will tell you what they saw when they played and it will be different than what the initial author saw. The game, in short order, will be other than what the originator intended. This is real collaboration and it’s terrifying.

Allowing your vision to be influenced by others requires an incredible amount of trust, especially from people who are trained to operate in an authoritative role. We (people like me, I mean) expect to command with a limited amount of dispute after a certain (brief) period of initial research. But when everyone at the table is at least as smart as you, you can no longer afford this ego-friendly attitude. And it makes you realise that, because the disputes and changes are emerging from real play and an effort to have real fun with diverse brains, it is a reflection of the interests of a potential audience as well. And consequently you are tricked into respecting the audience in the same fashion. The game now has its own design goals (or perhaps the design goal of the gestalt brain that the collaborating team represents) and these goals are roughly representative of the audience, because any gathering of four or five people probably has many general qualities that are represented by a much broader subset of people. I might not be well represented by any substantial group of other people, but the intersection of the interests of five of us certainly is.

So respect and trust are created by earnest iteration over play. Play. Let’s say that again. Play. That’s an order.

–BMurray

It’s time for me to think out loud about the game document as an application. All this talk about electronic books and manufacturing buckets and so forth has me thinking that practically every instance of the electronic book is at least as flawed as the paper book in some way or another. See, what the book does right is convey the content via efficient and effective use of the medium. What the electronic book does so far is attempt to mimic the book or extend it incrementally.

Seriously, an electronic device with gigabytes of space a millions of cycles per second and the best we can do is pretend to be a four dollar book? Fuck that.

Exploiting the new medium — a handy computer that does stuff besides display “Hello world” — is going to take some serious innovation. This is not a matter of making new standards or writing books using them or any of that. Real, serious changes to use the machine to do what we really want to do.

See, the game text is a compromise between what the author wants to do for the end user and what the medium is capable of doing. So when we make the machine pretend to be a book, we adopt the same compromise that the author has made for almost 700 years. I think maybe we can do better. Moore’s law, applied since 1450 or so, suggests we can do astronomically better.

What the author wants to do would, in my industry, be encapsulated in a requirements document. We don’t generally do this for books because the compromise we accept cuts so very deep — there’s just not all that much we can do. By contrast, when developing content for a computer, you want to start with the assumption that we can do anything. So now we have to constrain to what we want to do.

So what kinds of requirements would a role-playing-game-delivery-application (usually a book but now released from these bonds) have? Well, a place to start is a little use-case analysis. Here are our users (assuming a traditional RPG structure):

• all players
• non-GM players
• GM players
• readers
• reviewers

Yeah, see, even at this early stage in the analysis we already see that we have vast possibilities open to us just be acknowledging that we can be different things to different people. So let’s look at the most function-rich (I’m guessing!) user — the GM. What does he need this document to do?

• Display rules that are in the context of the current action at the table
• Search for a rule by keyword
• Display action-relevant information for each character
• Display action-relevant information for NPCs and monsters
• Adjust resource elements (fate points, gold pieces) for all players and maybe the environment
• Roll dice
• Display and modify campaign notes relevant to the action at the table and upcoming action
• Record table audio and video
• Convey private information to players (and observers?!)

I could go on. But starting with the assumption that we have a general purpose computer with audio-visual capabilities, a network, and some storage, we find the doors blown open on “what is an RPG if not a book?” As a GM I should be able to award fate points (in secret and in private), get updates based on player activity (“I am tagging ‘Zany funster’ for +2 because I’m just so awesome to be around.” click and fate point tallies on all machines are updated), see what aspects are begging for compels (maybe literally — a player might flag an aspect as a fun thing to tweak and the GM can respond to that red flag). It’s packed with back-channels that are both in and out of story.

So this is what I’m working on right now — what are the use cases for a Next Generation RPG Delivery System? Then after that will come the requirements proper. Then a design. And then I start developing iPad applications? Maybe.

I better go buy some books on that.

–BMurray

Okay, first an explanation.

Yes, there’s not been much here over the past couple of weeks but that’s not because I’ve stopped writing. I had a bout of sinusitis a couple of weeks ago that was crippling. In fact it was painful enough that I started researching the nature of Purgatory through the testimony of Saints. Anyway, this post isn’t going to be about pain or Purgatory or religion. In fact I just started writing so I’m not entirely sure what it’s about yet.

Last week I finally registered VSCA Publishing as a real company. I’d been dreading that almost as much as I was dreading doing taxes (which I left until the eve of April 30th) because I am really not interested in business. It’s a fun kind of game, but I’m just not a business guy. I don’t care if I make money and that seems to offend a lot of people, especially the government. Well maybe not especially the government, but especially the Chamber of Commerce and other similar entities where I might go for help. It’s like explaining calculus to a dog — I just want to make cool stuff and sell it for some money. I want lots of people to see it and I don’t want it to be a burden on the rest of my life to do it. If the net at the end of the day is zero (after buying all the cool tools I need to do the deed), I am so very happy. What you get is a look like a fish watching a shiny thing. Buh? Sha-wuh?

Anyway, what I finally did was go to a web site that does business registration in BC (recommend recommend recommend) and did the whole thing with a very friendly click-click-click process that led me through a questionnaire that hooked me up with all the forms I needed to fill out, guided me regarding the contents, charged me a reasonable fee, and spat out all the registration numbers I need. Done. Holy crap that was easy and I didn’t have to justify my choices to anyone.

So now VSCA Publishing is the real deal — a sole proprietorship owned by me and licensed for import and export and all that good stuff. Most of the acronyms on my new tax forms make sense to me now, so next year I won’t have to guess what numbers to put in them. I may even hire someone to guess for me.

With VSCA Publishing now an honest-to-God registered capitalist Entity, a de jure cog in the great engine, I will disclose the status of current projects.

Chimaera is front and center right now. JB Bell is the primary author on it and it needs some authoring done, but it’s had some great iterative play tests and the rules are coming together solidly. I’ve been doing cover art experiments and layout tests and am confident we have a path to tread here that will be exciting and fun and worth your money. I don’t know how long it will take because the real work is still ahead and while everyone loves to wonder out loud about game design and to test shit, real work is hard. We’ll see.

Soft Horizon is on the back burner because it needs a lot of play to get it to a place where I can confidently talk about producing it. It has a cover and a solid layout design ready, and there’s a lot of text in place. It could go to press in a month or two if not for the fact that it just hasn’t been played enough for me to be confident of its quality. And because Chimaera is our current focus, we won’t see a lot of play time for this for a while. That’s okay — I know from experience that taking a break from a project is a tool.

Hollowpoint is kind of ready and kind of not. It’s a weird game because it’s a major departure from more traditional structures — it’s more akin to Gregor Hutton’s 3:16 than Evil Hat’s Spirit of the Century. It’s a great energetic one-shot and it has good linkages for campaigns and I kind of adore it, but something nags me about it. I’m thinking I should just write it and get it out there for other people to judge. It’s weird and so I don’t want to show it, like a baby with a tail. But having a tail would be really cool. It has a cover and a layout design. I should sit down, gate it, and write it.

Soulscape is in a weird place. It’s Toph Marshall’s baby and I don’t know what I think about it. It’s like one of those cool looking toy puzzles that you just can’t figure out how to play with. It’s cool. It’s elegantly crafted. But what does it do? Is it fun? Is it art? Can I explain it to someone else? I’m not excited about this project but I am intrigued by it. It’s a game you might tag on to another game — you might play it to make sense of a D&D TPK. That’s a mighty narrow niche, but also kind of awesome. In fact, I may have to engineer a TPK in an upcoming game in order to playtest Soulscape. So I can’t say where this game is on the timeline yet because I don’t really know what it is yet. It is something though.

Finally, I am sad because my telescope is rotting in storage. Living in the city is great for so very many things, but not for being an amateur astronomer. The beast is a headache to get outside and I don’t drive so I’m not taking it further than the building courtyard and the light pollution is just too severe. I’ve seen all I can see with it in this space and so it’s in storage. My experience with storage is that it’s just a step en route to the garbage and that makes me unhappy. I can recover my costs by reselling the optics — that’s not an issue. I just wish I could live the life I want in a place where I can see the night sky.

–BMurray

No this is not about coping with fans that won’t shut up.

Any system over a certain minimum of complexity (and that minimum is pretty small) risks accidentally incorporating a positive feedback loop. There are three things that come out of this observation: first, how come? Second, so what? Third, holy crap what do I do?

The reason this happens is because eventually a system (and here I mean any system, not just game systems) can get to the point where you can’t hold the whole thing in your head at once. When that happens, it stops being clear what the follow-on effects of a function are because they can be indirect by several layers. Function A causes Function B to trip which incidentally calls Function C which needs Function D which creates an error condition for Function A in an unexpected context. That’s not actually a positive feedback loop but an example of a different kind of error that sometimes manifests as a “priority inversion”. We don’t care about priority inversions here.

But the point is that you can no longer rely on your knowledge of what affects what. Worse, if you have multiple people developing, you will probably only understand their area of participation to some minimum level. With games that’s the operational level — you have to really force yourself to care more deeply than “does it work at the table”. How it works and what the ramifications are can’t be your focus for every single function when something is past some minimum level of complexity.

Okay so one of the things you might not notice is a positive feedback loop. That’s where something increases some resource, which increases the thing that’s doing the increase. If there is no counter, something will balloon out of control. Or rather, as you didn’t notice it in playtesting, something can increase out of control in some case. Maybe an edge case or maybe just something you didn’t get around to testing.

Negative feedback loops, on the other hand, are a good thing. They are stabilizing: when A decreases B which decreases B’s ability to decrease A, the system damps back to a stable state. That’s good.

Now in my Real Job I care about positive feedback loops because they cost millions of dollars if they get to the field. So I do something called “feedback analysis” and it’s really simple and you can do it for game systems you’re working on easily enough. What you do is this:

For every resource in the system, draw a bubble and write the name of the resource. So for Chimaera right now I have Desperation, PC successes, Enemy successes, PC health, Conversion, and Take. Desperation is a pool of dice, successes are the number of successes you might get in a conflict round, conversion is the rate at which you can change Desperation to harm, and Take is the rate at which you can steal Desperation dice to roll for yourself. You can see that you have to treat “resources” as a fairly abstract concept in order to make this work.

Okay now, for each resource, consider whether or not its increase affects each other resource. If its increase increases another resource, drawn an arrow from source to destination and draw a + on the arrow. For example, Desperation increases cause Conversions to increase. Then, again for each resource, consider whether or not its increase decreases each other resource. If this is true, again draw an arrow from source to destination and draw a – on the arrow. For example, Takes decrease Desperation.

Once you have your map, you should have a lot of bubbles with crisscrossing arrows. Now you are ready to analyze.

For each resource, follow a + arrow out. Then follow a + arrow from that one to another. If you can get from a resource back to itself by following + arrows, then you have a positive feedback loop. This is where your game can explode.

Positive feedback loops commonly manifest as “death spirals” which you sometimes want, so if you find one look to see if it’s damped by some negative inputs. There might be enough to keep it fun — but watch that resource in play. That’s a red flag zone for you.

Sometimes positive feedback loops are actually loopholes in character creation that allow unexpectedly powerful characters. Or unexpectedly crappy ones (positive just means a resource is increasing — if the resource is something that sucks, then it’s increasing the suck).

Most often positive feedback loops result in optimal tactics — things that there is no reason not to do in a tactical scene. These are a pain in the ass because if you find them late, sometimes all you can do is shrug and say, “the GM shouldn’t allow you to be a douche like that”. This is a pain because, although this is true, this is how we patch games with serious defects so that they work. Obviously it would be better if the game did not have those defects to patch.

So there’s a tool for you design toolbox. You’ll find it’s something of an art when applied to role-playing games because the resources are not well-defined. That’s cool. I like art.

–BMurray

Designing games is fun, but having several game designs on the burner has the uncomfortable side-effect of making every game session also a design session. And I kind of miss just gaming.

When you’re working through a design by play, which is basically how we roll, the flow is constantly interrupted with discussion — theory, effects, probability, and so on. It’s very hard to get at the emergent story that we play for in the first place, because we are so immersed in the mechanism. If a scene goes into unfun territory for mechanical reasons, it behooves us to figure out why, re-jigger the mechanism, and re-run the scene to make it work. The end result is that these sessions are very technical — they are sequences of mechanical application with only the thinnest material joining them, because we already know that that thin material can get spun out to story in a real game. When you’re “just gaming”.

While building Diaspora we did get to the “just gaming” point, where we had the system smoothed out enough that we could really play and build stories and have a great time gaming, so maybe I’m just frustrated at having a bunch of games in the mechanical stage. On the other hand, when I tried to get 4e working, or even 3:16 for that matter, I had that same jointed mechanism feeling — hopping from one mechanically guided scene to the next with interstitial role-playing that was perfunctory or absent. So there is also a risk that we’re making games that aren’t what I like playing, because if that’s all we can do then I’d rather play something that’s about the mechanism. Something that doesn’t even pretend to have a more creative context. Settlers of Cataan, say, or Wings of War.

So I think my objective for the next few weeks of gaming is to find the story and I think I know where it’s hidden. And if we can’t find the story then we need to seriously re-think the new designs. My suspicion is that what’s missing is compelling non-player characters — there’s not enough conversation with people I care about on topics I care about. Or rather there is but it’s with the people and not the characters. Those conversations turn the interstitial hinges joining mechanical play into the memorable part of the evening. This relegates the mechanism to the space it should be. In playing “properly” it may serve to highlight mechanical issues from a more interesting angle, too.

I think this is born out by our two Burning Wheel sessions from last month. The highlights of play, for me, where a couple of conversations with unstatted NPCs. They utterly lacked any mechanism (though the entrance points and exit points were mechanical — it’s certain that mechanism let the NPCs be awesome by setting the stage and supported the exit) but they are the images I have now of “what was fun”. The ghost of an ancient orc king, from back when orcs were artisans in their hate and not just monsters. His speech let me create a whole culture right there for the players, and that kind of thing is why I game.

So maybe it’s just time to start thinking about gaming as a hobby again. Maybe the business of it is too close; too looming. I’m not really interested in business. Or maybe I was but now I’m bored of it. We need to play.

–BMurray

For Chimaera, I think maybe everything is an opposed check.

A static check is when the player rolls her character’s dice to beat a fixed value. The GM sets this value (call it an obstacle) based on her conception of the problem to be solved.

An opposed check is when the player rolls her character’s dice as normal, but the GM rolls opposition dice that reflect the likely degree of difficulty in the obstacle — that is, her conception of the obstacle is the expected result of the roll. It could go a different way, however.

The critical difference between these two is the weighting of invention and discovery. My own preference (and clearly everyone has to state their preference on these things so that we can work our way to the holy grail of The Perfect Game) is “it depends”. Seriously, both are cool. But I want to talk about what kind of cool the opposed check is.

The opposed check emphasizes invention. Now in both systems there is a certain amount of invention — when the player gets an unexpected result (unlikely success or failure) it behooves us to narrate what happened. We invent some new story to explain the result. This is fertile ground — some of my most memorable gaming stories are about unexpected rolls and the subsequent narration invented to explain it.

The opposed check, however, invites narration about the obstacle as well as the actor. And in this we get something new: we get to apportion blame, which is always fun. So there is a second possible form of invention here and we just saw that invention based on the dice is often memorable. And fun.

Let’s say, for example, that the task is heart surgery. The surgeon is doing a routine valve replacement. The GM invents the obstacle — she has decided that this is a routine valve replacement and sets the obstacle value at 2. The player rolls her dice to beat 2. She does or she doesn’t, but the space to explain the failure is solely at the character: why did she fail at such a routine operation? She must have fucked up, rolling a zero like that.

If, however, the GM sets the obstacle at 4 dice, which have (let’s say) and expected value of 2 successes, then this is the same surgery. Let’s say that the surgeon gets 3 successes which would handily beet the fixed obstacle. The GM, however, rolls well — 4 successes! — and the operation goes to hell. The patient dies. What happened?

Happily, we have a new axis of information from which to invent the narrative for the failure! We know that the surgeon did very well, actually — 3 successes. We also know that the difficulty was much higher than expected — an obstacle of 4. So we can deduce that the failure was due to something unforseen about the problem itself rather than a failure in the actor. She went in to do a routine valve replacement and discovered a second valve in much worse condition that was masked in diagnosis by the noise of the first. The second valve tears during the routine repair of the first, the patient inexplicably loses blood pressure as the torso fills with arterial blood from an unknown source, and the beep machines makes its lethal squeal. The patient dies and there’s nothing the doctor could have done. In fact she came close — 3 successes to an obstacle of 4 — and did everything in her power to solve an almost impossible problem.

This is what’s rich about having to invent based on the dice and you double the richness by rolling dice on both sides of the problem. The downside is, of course, the cost of invention. This puts more onus on the players to fabricate the results — to interpret the dice — and this is not necessarily to everyone’s taste. And not all the time. But if the game is intended to drive improvisational explanation, this is a powerful tool in the kit.

–BMurray

In the last post we opened a can of worms about specialization and how it relates to the latest iteration of the Chimaera character design system. I had some ideas about this on the bus and I have a long bus ride, so by the time I got to my desk they were more post than comment. So here it is.

So the issue is this: in the current system, a character specifies Paths which have skill ranks that will be checked in a task. She can bring in Distinctions, which are phrases attached to Paths, for advantage, but only if they are associated with a different Path than the one being checked. This means that in order to get bonuses to your Soldier Path you need to be good at it because of things in your other Paths, like, say, Farmer or Hunter. Some say this makes it unsatisfying to specialize. I will show that it is not.

The most boring possible specialization is the Neurosurgeon. Here is a guy who has spent his entire life learning to do one thing extraordinarily well. To represent him in a classic hierarchical skill specialization system, he might look something like:

Doctor
- Surgeon
- Neurosurgeon

…and, just so he can’t also be a first string NFL linebacker, we assume that each nested level of specialization is more expensive than the previous. This is so that we can model the dedication and expense of becoming a neurosurgeon: there just isn’t room in this guy’s life to also be a linebacker.

Now the problem with this specialist is obvious: he’s extremely boring. I will suggest that this is because his story is not captured in his skill set and I want the story to be captured in the skill set. Neurosurgeons are not, by and large, boring people. They are often fascinating people despite having dedicated themselves to a single highly specific purpose. So let’s look at this specialist in Chimaera terms and try to find a way for this guy to get 8 dice into a neurosurgery conflict without stretching any Distinctions too hard. 8 dice would be an awesome number to roll.

Well first and foremost he’s a surgeon.

4- SURGEON

• First, do no harm
• Rock steady hands
• Calm
• Assertive

I’m trying to find Distinctions here that are not expected to be used for surgery checks but that also are part of being a surgeon. This is because I can’t tick them when doing surgery — so these are the ancillary features of being a surgeon. Okay now let’s specialize him. I think he worked his way through college as an EMT:

1 – EMT

• Get it right the first time.

That’s it. It’s not his dedication and he might use this Path when performing emergency first aid or something like that where SURGEON is not appropriate. But we can see how this is a no-brainer tick for SURGEON. Okay now I think in order to be such an extraordinary specialist he is also a noted researcher in his field.

2 – RESEARCHER

• Expert on the brain
• Meticulously logical

Okay, again we can see how these might come into play in surgery and in particular in neurosurgery. Now this guy is a brain expert. Says so right there. Finally, I want to give him a hobby to flesh him out. I like making hobbies the next most important thing after the top Path because that’s how most interesting people are in my experience. So I think he has a hobby of watch repair, which I will generalize as TINKER.

3 – TINKER

• Take anything apart
• Put anything back together
• Keep perfect notes

Okay, that’s my specialist. Now let’s say he’s deep in a very difficult brain surgery — which he specializes in. He has 4 dice for SURGEON and can call on “Put anything back together” and “Keep perfect notes” (from TINKER) without controversial justification. He can obviously tag “Expert on the brain” from RESEARCHER and “Get it right the first time” from EMT. Without too much more work he could tick “Take anything apart” from TINKER as well I think. So in a really hard surgery (brain surgery mind you) he could call on 9 dice — extraordinary capability! — and the result would be that he is Exhausted (two Paths completely ticked). Okay, so the specialization works to make a guy who can do Neurosurgery.

But look at this guy:

4- SURGEON

• First, do no harm
• Rock steady hands
• Calm
• Assertive

3 – TINKER

• Take anything apart
• Put anything back together
• Keep perfect notes

2 – RESEARCHER

• Expert on the brain
• Meticulously logical

1 – EMT

• Get it right the first time.

He has an interesting past and an interesting present. He can easily get 5 or 6 dice to make or fix gadgets. He can probably find 4 or 5 dice just to convince people of stuff without even calling on a skill — just a 0 rank Path and a bunch of ticks. And he has a few things in there that are very plausible hooks for compels, assuming Chimaera has compels (and for JB I’ll say, yes it does: the GM can compel any ticked Distinction and the payment is unticking it — elegant and no new bookkeeping).

So that’s my argument for specialization in this everted hierarchy. In order to specialize, do what real people do: dedicate your past to your specialization. Now you are specialized and also risk having an interesting past. And present.

This is way cooler to me than Doctor -> Surgeon -> Neurosurgeon.

–BMurray

Right now the VSCA has several games percolating. It’s interesting how they swim in and out of focus.

Chimaera is the latest example. Until recently it was mostly still in JB’s head and it seemed to derive largely from fiction he wanted to tell, and consequently lacked a presence as a game. Instead it was more a sequence of visual ideas, which is fertile soil for a game to germinate. But it’s not a game. Unfortunately, it wasn’t a sequence of visual ideas that was all that compelling to me and so I was having a hard time assisting in its transition from “cool stuff” to “game”. It was out of focus and I honestly saw no game there (though I did see a number of cool ideas that could become part of other games).

On Easter Friday we sat down and went through community and character generation for Chimaera, something that had a few rules written down but nothing as concrete as a checklist. Now part of the problem with designing in your head is that the order of use makes you go in circles. When you sit down and do it, you can iterate in those circles, though, and design and pare and sculpt and fix and text and again like that. So at the table instead of one linear idea, we got this happening:

“Here’s how you make communities.”

“Hrm, I don’t have a feel for my community with these numbers. Too many zeroes. It’s, what, ‘disorganized’ and ‘isolated’?”

“Maybe a different curve? Fewer zeroes?”

“How about this? No. This? No. This? Whoa that’s cool; low probability of zero, peak probability of one, and descending probabilities to five.”

“Zero should be special then. Not ‘least’ but maybe opposite or transitional. A new state.”

“Try it first.”

“Okay my community seems cooler. But what’s Integrity zero now? Just ‘chaotic’ or ‘unstable’?”

“Should be a new state. It’s not a community at all at Integrity zero — it’s property for someone else. Call it ‘commodified’.”

“So my community is a baby farm that your community is using as a buffer to keep the daemons off? Holy crap that’s dark.”

“And cool.”

And like that. We built communities, tweaked the rules, built them again, tweaked the dice, built again, and so on until we found a place that wasn’t “good enough” but rather “awesome”.

Now with characters you have something more complex happening and this makes design necessarily iterative for the same reasons that it’s hard to figure out what order to make the chapters in the game book: in order to make a cool character you need to know something about the resolution system so that you can see how the character representation will impact play. In order to test (or demonstrate) the resolution system, though, you need some characters.

So JB already had a character generation system established but it wasn’t very clear exactly how the resolution system would get pushed around by the character representation. So we went straight to the resolution system and did something that I will suggest to any designer: quit dicking with the dice. It’s not that important. Pick a dice system and then make the character representation play the dice like a fiddle.

So the dice system we went with is simple: roll a pool of dice based on something. Odd numbers are successes. Most successes wins. For now, who cares how you get the dice pool size.

We rolled some opposed checks and it’s good. It creates successes and failures and ties. Taking a card from Hollowpoint’s deck, we decided that ties are mutual failures mostly because that’s not boring (a wash is boring) and because it speeds up rather than slows down the progress of a multi-stage conflict. So if you lose, you lose as many dice as the difference between your and your opposition’s roll. If you tie, you both lose a die.

Now here’s where cool dice mechanisms often fail to meed the road with any grip: how do three players conflict against opposition run by one player? There are two basic choices: take turns and resolve piecemeal, or all roll at once and discover. We’re all a little tired of you-go-I-go-you-go systems and dice pools are amenable to roll-and-discover so it goes like this:

Each player rolls her pool. We test with sketched characters that have skills and aspects, though we don’t really know what these will mean yet. Ranking the skills, players have four dice in their best skill. See what happens.

The GM rolls an opposition pool that’s big — say four dice per opposing player for an equal match.

The GM can use successes in her pool to cancel successes in any or many players’ pool. Shades of ORE and Hollowpoint. If a player’s pool has been reduced to zero, the GM may declare the rest of her successes as damage on that player. Or she can use them to cancel more successes in other pools. For each damage unit, the player loses one die somehow.

Players with remaining successes can use them to damage the opposition. For each success the GM removes a die from her pool.

Cycling through this a few times it seems to work. It’s fast, it’s exciting, it’s a little tactical, and it’s easy. It always has a resolution.

So we go back to character representation. Characters have Traits which are like skills or something. They are ranked from four to one. JB previously had the idea that each Trait has a number of Aspects equal to its rank. That’s cool and simple. And it’s suggests a mechanism that it turns out binds all aspects of resolution together in a tight little bundle:

You choose the Trait to play with and get that many dice. You may tick off one or more Aspects under any other Trait and add a die for each. When you take damage, you must tick off an unticked Aspect. You can’t use a ticked Aspect. If you have no unticked Aspects, you are dead.

Okay so now you have a way to record damage, a way that damage impacts effectiveness, a risk/reward balance for grabbing extra dice, and something cool implied in character generation (here we iterate): you do not want Aspects associated with a Trait to be about the things that Trait is obviously about, because they cannot be used to assist that Trait! We smell something cool.

See, let’s say you have this character:

SOLDIER 4

• “Crack shot”
• “Martial artist”
• “Marksman”
• “Ruthless killer”

FARMER 3

• “Grows great corn”
• “Weatherbeaten”
• “Animal handler”

Now, in a fight she will use SOLDIER and grab four dice. But what will she draw on for additional effect? “Grows great corn”? So it’s mechanically defective but, worse, it’s boring. Very little under SOLDIER tells me anything interesting. However, a player playing tactically during character generation (and I feel strongly that a game ought to do best when played best) would choose Aspects that are not strictly related to the Trait in order to maximize usage in other contexts. Say:

SOLDIER 4

• “Born leader”
• “Can march for days”
• “Plans everything”
• “Cool under pressure”

FARMER 3

• “Varmint hunter”
• “Strong as an ox”
• “Handy with all manner of dangerous tools”

Okay now we have two valuable new results: first we have a character that will play well at the table. When making SOLDIER checks, she can grab “Varmint hunter” for shooting and the other two for hand-to-hand. We just learned that because of her background, she’s a soldier with powerful close-combat specialties. When making FARMER checks, she’s grabbing her planning skills and her endurance and even her leadership for successes, and so we learn how life as a soldier impacts her duties as a farmer. So the second valuable result is that this character is way way more interesting!

So we run a fight again with these new characters, and we find that there is benefit in switching out skills (a common failure in games is using the same skill over and over which is boring as hell)! If you use a Trait and tick an Aspect or two, it will often pay to use the Trait with a ticked Aspect next because the skill you used last time has lots of Aspects available! And, further, damage will influence your next choice as well. This means that your best Trait is often the optimal choice initially, but as the fight evolves it may become the worst choice because it has all your unticked Aspects under it. This is really cool — combat becomes a series of dynamic choices. It’s tactical, and the tactical choices drive the creation of unexpected stories. How can I get my FARMER Trait into this desperate battle with mutated wolves?

Any really great session winds up generating art. So, in celebration of the fact that Chimaera began as a set of visual ideas without cohesion (well, without game-like cohesion) and finished the evening as a concrete and delightful game, I went to the drawing board and worked up an image to incorporate into a cover for the game. One that has none of these visual ideas. I picture it in the upper right corner of a field of black. Because I smell a very dark game here.

–BMurray