Ars Magica Video Game Kickstarter

We (Atlas and I) are working with Black Chicken Studios to produce an Ars Magica video game. (I suspect that Black Chicken will be doing most of the work, being the video game people, but we will have a lot of input as well.) Black Chicken have launched a Kickstarter to fund it. Atlas has a press release about the game and Kickstarter as well.

The target for the Kickstarter is high, because it has been properly budgeted, and $290,000 is actually enough to make sure the game happens. If there isn’t enough interest to pay for the project, we won’t do it, and thus avoid losing $290,000, but we really hope there is enough interest. Is there any interest here?

Choosing the World

Although I will, of course, develop the detailed design of the game world while I’m working on the game, I do need to choose the basic type of world I want to create. Since we are focusing on things I’m personally interested in, there are four options: “classic” fantasy, modern fantasy, science fiction, and historical.

By “classic” fantasy I mean, essentially, the sort of fantasy world popularised by Tolkien and his imitators. Elves, dwarfs, orcs, swords and sorcery. That sort of thing. Yes, a lot of role-playing games have used this background, most famously Dungeons & Dragons. However, the basic aim of my game is sufficiently different from that of D&D that using a familiar background might actually be a good thing, if I’m aiming at an audience of role-players. If the background is classic fantasy, the players’ assumptions about the general ways in which the background works will be correct, which will help them get into the game, and make sensible choices from the background options available. As the basic goals of the characters are very different from those in other role-playing games, a bit of familiarity could be a big help. In addition, this is something I keep coming back to when I look at classic fantasy worlds; I really like them, no doubt in part because they were an important element in my formative years, but the games set there do not support the sorts of characters I actually want to play.

What about the other sort of fantasy, modern fantasy? If I wrote a modern fantasy game, I would probably set it in Japan. It would be something like Tamao: The RPG. Of course, that would make the background very unfamiliar to the players, apart from the handful of English-speaking role-players who live in Japan, unless I wrote it in Japanese and targeted the Japanese market. However, I suspect that’s a bit too ambitious for the moment. I could set it in the UK or the USA, but that removes one of the big advantages of putting it in Japan: I can’t use information and images that I can gather just by walking around the area where I live.

Moving away from fantasy, if I created a science fiction world it would be fairly hard science fiction, not space opera. However, since the things that characters would create would include new scientific theories, it couldn’t be hard science fiction according to the strict definition, because the characters would discover things that broke the currently-established laws of nature. (This is, however, a fundamental problem for hard science fiction; new scientific laws will be discovered in the future, so if you set something far enough in the future it’s not hard science fiction if you don’t make something up, but also not if you do. Obviously, there are ways round this, but it does make the standard definition less applicable than it might be.) I’d probably set it in a somewhat transhumanist setting, like those used in Transhuman Space and Eclipse Phase, because those settings provide a lot of space for engineering creations as well as for fundamental scientific discoveries, and thus make it easier to create a setting that supports a long-running campaign. It’s a bit hard to maintain continuity, and interest, if you completely restructure the world every week.

Finally, a historical setting would probably start with a straight historical setting, without fantasy, but the actions of the player characters would quickly turn it into an alternate history setting. They might make scientific discoveries and change the technological background, or create important new works of art, or change the political structure. In any case, the aim of the game would be to enable the player characters to change history, in deep and fundamental ways. However, the first changes might be quite minor, so this would also have an educational aspect, as the players would naturally learn about the historical period while playing.

The downside of both a hard scientific setting and a historical setting is that a large amount of research would be needed to do them properly. I could probably do a historical game set in medieval England based on the research I’ve already done, and I have enough scientific background to be able to make a stab at the science fiction, but in either case I’d have to do even more research. This isn’t entirely a bad thing, because I like research, but it is another barrier to getting started on writing the actual game. I probably don’t need any other preparatory projects to waste my time on.

To a lesser extent, the same can be said of the modern fantasy setting. While I only need to go outside and walk around for some things, for others I really would need to find out how Japanese society works in that respect, and that might not be at all easy. Again, it’s something I want to know, but research takes me even longer in Japanese than in English (my Japanese reading speed is still not the same as my English reading speed), so it would, again, significantly delay the creation of the actual game.

Looking at all of these issues, it rather looks as though the classic fantasy setting is the best choice to start with. I can just make the background up to fit the structure of the game, which is likely to be helpful while I am still trying to get things working. More generally, I can just make bits of background up as I need them, and, as long as I keep track of them, they can’t be wrong. I might decide I want to make revisions later, and there will be time for that, but I can’t make actual, unambiguous mistakes.

The next step, then, is to start actually writing the game. This process, however, will not lend itself to regular blog posts, and virtually nobody is reading these posts anyway, so while I think I will post about it from time to time, when there is something significant to say, this will be the last of this semi-regular series.

Setting the Scene

Introducing the game world is, like character creation, a major problem for role-playing games. Most games end up with several volumes of world information, running to thousands of pages; Ars Magica is certainly no exception to that. Reading this information and discovering the game world is part of the fun for a lot of people. However, feeling that you need to read several thousand pages before you can play a game does tend to discourage casual players, and make it hard to try new games. It’s even more of a problem for my design, where I want people to be able to start playing within fifteen minutes, and to do so without a gamemaster, and thus without someone assumed to know what is going on in the game world.

A bad approach to this problem would be to provide lots of material to be read out by the players. Reading is fun, but it isn’t playing, so this also delays people getting into the game. In addition, the players are not going to remember everything that is read out to them, particularly if it’s about an entirely fictional world, or something historical or from a remote culture. Boredom and confusion are the likely result, and neither of those is fun.

On the other hand, the information does need to be contained in the scenario, because there’s nowhere else to put it. However, it is better if the information is supplied as needed, rather than in large chunks, and if the information supplied is closely involved in the action being taken. This is the “show, don’t tell” rule for writing fiction, transferred to a role-playing context.

The concrete application of this is to the description of the actions the characters take, and to the description of the characters themselves. Each individual problem needs to be described, and that description should introduce part of the world. Further, each action that the characters take should be described, and further the world’s description. Describing the problem is relatively easy; this can be a bit of text that one player reads out. It can be kept short, and it’s immediately relevant to play, so boredom and confusion can be avoided.

Character actions are a bit more difficult, because the players should have control, but don’t have the necessary knowledge. I think the best approach may be to provide two or three options for actions in each situation, with a description, and then let the player choose which one the character does.

Character features, which grant additional dice to roll, should also describe part of the background. They won’t go into detail, but they’ll start to reveal something of what the game world is like. And, if they’re chosen, they’ll be linked to the character, and thus easy to remember and important. Of course, they should continue to be important in later parts of the scenario, but that’s just a matter of careful design.

Finally, there’s the background itself. What is the world like beyond the immediate concerns of the characters? Here, I would like to introduce elements that characterise the world, and the particular location where the characters are working. One possibility is to make two elements available at each action, with a brief description of each one. Then, the player can choose to incorporate one of the available elements into the description of her character’s action. If she does so, she gets an extra die to roll, and that element is confirmed as being present in the immediate environment.

A possible development of this is to have more detailed descriptions of certain elements, which become available when a basic element is in play. Later players can choose to involve them, making the description richer, and getting extra dice for their characters’ actions.

However, there’s going to be a lot to keep track of here, and scratch paper may not be the best way to do it. Instead, I think it might be a good idea to print the descriptions on cards, around business-card size. The cards can then be placed between the players, so that they can see what elements there are in the environment, and how they are built up. The cards could also have pictures on, although, as I can’t draw, that is unlikely to happen in the first stages of development. Actually, the whole thing could be made multimedia, with images and sounds, even animations, and then run on a laptop, tablet computer, or smartphone. However, I’d have to learn to program to do that, as well as get the multimedia from somewhere, so that’s also not going to be the immediate form this takes.

Another advantage of using relatively small cards is that it limits the length of the descriptions, which is an important consideration. The elements might build up into a lengthy, detailed description, but they start small. Since adding detail means going back to an element that has already been added, that will remind the players of what is around them, and, again, makes it important that the individual elements are short. Reading over a lot of long descriptions repeatedly risks creating boredom again.

This is going to require quite a lot of work from the author of the scenario. He will have to make sure that the elements work together well, support the story of the scenario, and let the players generate interesting characters. They will also need to be structured so that it is very difficult to produce something that doesn’t make sense. Thus, each background element card will need instructions for what replaces it in the pool of available elements if it is used, or, for incompatible detailed descriptions, what has to be taken out. For character elements and elements of the thing being created, the scenario needs to specify what is available at each point, possibly depending on what was chosen earlier.

However, it is important to avoid closing off options. We must remember that the players do not know what they are doing; they should not be allowed to make choices that mean they cannot do, or be, something significant in the game world, at least not early on. For example, in a classic fantasy game, players should not be able to exclude the possibility that their character belongs to a particular race until they know enough about the races to make that choice sensibly.

Again, looking away from the scenarios, this information can also be gathered into books about the game world, so that they are available for people who like to read that sort of thing. A player could learn about the background from books, which would just mean that he knew more about how the elements introduced in a scenario connected to the rest of the world. That wouldn’t, normally, be an advantage, since the characters would be assumed to know even if the players didn’t. It would also be possible to produce packs of background elements, for use with any scenario, that introduced new features of the game world.

I’ve been very vague in this as to the nature of the background, because that’s the next thing I have to decide. In broad terms, what is going to be the setting of my game?

Revising Creation

My suggestion in the last post that we could generate characters during play created the problem that it would work better if the basic rules relied on simply beating a difficulty, not on the amount by which you beat it. However, the rules I designed for creating something relied on the final total. Can I redesign them to use fixed difficulties?

We can start with the simplest level again. To make the creation “work” in a certain way, you need to beat a certain difficulty. If we keep the pattern of having about three aspects to a creation, which seems reasonable, that would require about three rolls for a whole project, which is obviously insufficient.

So, we start adding complications. If you try and fail, the resistance of that aspect increases by a certain amount, the same no matter what the failing total is, adding to the difficulty of future attempts. If you try and succeed, the resistance of the other aspects increases by a certain amount, again the same no matter what the succeeding total is, meaning that the order in which you approach things matters.

We can then add rolls to reduce the resistance, giving them fixed difficulties again, in return for a fixed reduction in the resistance. The rolls to get additional dice for your pool work perfectly well as they are, as long as we only include one level of success. That’s fine; more complex things can be introduced later in the game, when scores have been fixed and people are more confident about their characters.

So, will this work? We want the characters to succeed at just about everything, because a failure fixes features of their character, and we want to avoid fixing things this early. So, if there are three qualities in an encounter, that’s three rolls, one for each quality, to start with. If each success adds the same amount of resistance to the other qualities, and we are going to recommend reducing the difficulty to the base level, that’s another three rolls; one to reduce the resistance of the second quality, and two to reduce that of the third. Six rolls is, at least, enough for one each in any practical group, but it’s still a bit low.

We can also introduce rolls to get additional dice to roll, or give the option to purchase character features that grant additional dice in a limited context. This would help to keep the level of commitment to talents down, by allowing the characters to explain some of their pool dice through a feature instead.

Thinking about the big picture, shorter encounters might actually be better in this situation. We do want to finish creating the character, or at least the relevant aspects of the character, by the end of the scenario, so we probably want a relatively large number of short encounters, to give the players a good idea of what the various talents and abilities do. In these encounters, the difficulties should be set to encourage easy success, so that the characters who do act only have very low minimum scores in the relevant statistics. A difficulty of 4, requiring a 5 or higher to succeed, might be reasonable. Success while keeping a single die is certainly possible, but players are likely to need to roll more than one, which provides space to introduce features. A difficulty of 3 might even be usable, because there’s still a 50% chance that players will need to roll more than one die.

Let’s look at a semi-concrete example. Suppose there are three tasks, each of which has a difficulty of 3. Failing at a task adds 3 to its resistance, but failing is strongly discouraged. Succeeding at a task adds 3 to the resistance of the other tasks, but there are things you can do to reduce those resistances. These tasks also have difficulties of 3, and reduce the resistance by 3. For each task, there is a character feature available that adds one die to roll, and an action that another character can take, with a difficulty of 3, to similarly add a die. We’re going to need features available to make more dice available for the rolls for assistance, as well, because we don’t want them to push the defined statistics up too high. However, support tasks for the support tasks are a bad idea in general; things will get rather recursive.

So, once we have gone through the bits of the setting and rules that are relevant, we can move on to more difficult tasks, where players can choose what they want their characters to be good at, and start to define them. The short encounters would naturally be part of a larger problem, but it’s probably best not to provide any major choices at this point, as the players don’t know what they’re doing. However, the more difficult tasks in the second part should allow for real decisions about strategy, and the first part of the adventure should have provided enough information for the players to make such decisions sensibly.

The biggest remaining question is how to justify the fixed difficulties within the game. It doesn’t make much sense for there to be thresholds of success in writing a novel, for example. You can write a readable and entertaining novel, and still have plenty of room for improvement. (Been there, done that.) However, it does make sense in what might be called engineering problems. For certain problems, you can solve the problem, and it really makes no difference how much extra brilliance you put into it; the solution won’t be any better. Obviously, there are exceptions, and the game should cover them, because they are the great technological breakthroughs that revolutionise society, but the more mundane problems do exist, and so they can be used in an introductory game.

Of course, this means throwing away my example pretty much in its entirety, but that’s OK. The example was just to help with clarifying the basic structure, rather than a serious suggestion for what I was going to use.

At the end of the introductory scenario, the players will have spent a certain number of points, and defined a number of aspects of their characters. Note, however, that you could also create a character just by spending the points, and such a character would be perfectly balanced with one created in play. This means that I can also produce a big book full of character creation options with point costs, for people who like doing that sort of thing, without biasing things in favour of people with the book. People who create a character in play are, after all, guaranteed to have lots of statistics that are useful in play, because something is only defined when it is useful in play.

In addition, this change to the basic rules still doesn’t require a gamemaster. The scenario needs to define the difficulties, but the players and characters are supposed to be aware of them, so there’s no problem with defining them out in the open. However, there is a problem for the introductory scenario. Someone needs to describe the world, but there might be no-one at the table who knows much about it. How can we approach that problem? That will be the topic of the next post.

Creating Characters

Character creation is an essential part of any role-playing game, even if it just consists of choosing a character from a list. Indeed, it’s a part that a lot of role-players really enjoy, including me; I’ve created quite a lot of characters that I’ve never played, and knew I never had any chance of playing. However, it takes time. For a lot of role-playing games, it takes over an hour, and that’s after you’ve read and understood a rulebook over 200 pages long. This is a major barrier to getting people involved in the games, and to trying new ones.

On the other hand, there are games with really, really simple character creation. The problem with that is that the games don’t support much differentiation between characters, or, indeed, the fun of designing a character. (It is a mistake to assume that the only fun part of role-playing games is actually playing them.) Thus, people tend to become unhappy with simple character generation systems after a while; witness the tendency for games to include more character creation options in later supplements.

Since I like creating characters, I want to make a system that permits that. However, I also want people to be able to start playing within fifteen minutes of sitting down with the game. I don’t think this is actually impossible.

My proposed solution is to have the characters created in play. That is, you start playing without a character sheet. You might start with a name, but you might not. You certainly don’t start with anything as definite as a concept, because with only fifteen minutes of preparation time you don’t know enough about the game setting to be able to come up with a concrete idea as to the sort of character you want to play. Instead, you start knowing that you are working with the other player characters on this project, but you don’t know why. That’s something you’ll develop during the game.

So, how is this going to work? Let’s look at talents and abilities first.

First, the scenario presents a problem, which can be addressed with a particular talent and ability combination. One of players decides that she would like her character to tackle that problem, and rolls a single die. If she likes that result, she stops. On the other hand, she might well want to go for a better result, in which case she rolls additional dice, one at a time, until she gets a result she is happy with. Once she is happy with the result, she gives her character the talent and ability required by the final result. The talent will be equal to the number of dice rolled, while the ability can be varied, depending on how many dice she keeps.

Of course, there needs to be a mechanism to stop players just rolling and keeping a hundred dice for everything. So, we’ll give each player a pool of points she can use to buy talents and abilities. I suspect that talents should be more expensive than abilities, because they will be more widely applicable, so let’s say 2 points for 1 point of talent, and 1 point for 1 point of ability. This might turn out to be wrong; in the example developed earlier, abilities were more useful than talents, because there were about the same number of both, and dice kept improve your results more than dice rolled.

For example, suppose that our example player is trying to come up with an original idea for a novel. The scenario tells her that she needs to roll her Creativity and keep either knowledge of the subject matter or knowledge of a related genre. Of course, she doesn’t have either of those yet. The scenario also tells her that higher is better, and that she wants a result of at least 9. She decides to go for knowledge of the subject matter as her ability, and makes a note. She rolls her first die, and gets a 1. Not a great start. The second die comes up with a 2, so she keeps going and gets a 4. That’s still not enough, so she rolls a fourth die, and gets a 6. Stopping is an option now, because she can get over 9. She rolled four dice, so she must have a Creativity of 4. That costs 8 points. She could choose to keep just two dice, for a total of 10, and a total cost of 10 points, but she could also choose to keep all four dice, for a total of 13, and a total cost of 12 points.

A similar mechanism can be used for items or features that grant additional dice. The scenario can make them available as an option, with a cost attached, and players who want them can buy them for their characters, and then use them. This can be used to introduce significant elements of character background in play, and thus flesh out the characters. If things like sex, race, and age are going to be significant, the scenario should introduce them early on, because it’s hard to play for long without knowing even that about your own character.

What happens when more than one player wants to take an action? One possibility is just to agree on who will do it. However, it’s best to have a mechanism for when people can’t agree. So, I think that each player will start with a pool of points, about six, with which they can bid for actions. If a player chooses not to take an action, the pool grows by one. Players who want to take the action, or acquire a characteristic, can bid against each other. Only the winner has to spend the points from her pool, but if you chose to bid you don’t get an extra point.

For things that won’t come up again later, such as background features, it would be a good idea to have a mechanism that allows players to insist on having that feature for their character. Thus, even if you lose the auction, you can have a feature by paying an additional point for it. The other player still gets to take this action, however.

This structure will encourage turn-taking, while still allowing some flexibility.

However, there is a problem with this structure. A player has to decide on how good the character’s talents and abilities are very early in the game, before she really knows which are most important, or what she wants her character to be. Recall that the player sets the character’s talent and ability by deciding how many dice to roll and keep; there is no ambiguity about the actual level of the scores. Something a bit more flexible would be better.

This can be done if the rolls result in simple success or failure. If it doesn’t matter by how much a roll beats the difficulty, then you can say that your successful character has at least a certain level in the talent and ability, but may be better. If we look at the example again, this character would have at least 4 in Creativity and at least 2 in knowledge of the genre, but might have higher scores in both; that’s something that the player could decide later, when she knew more about the game, and her character.

However, in order to make that possible, we need to seriously revise the mechanics for creation. Can we do that? I’ll look at that in the next post.

Creating Projects

So, how do we stitch creative actions together into a story? Here, I think I do want to model things on the classic RPG pattern. It’s classic for a reason, after all. The pattern is to have a number of encounters, which go together to form an adventure. Adventures are then strung together into campaigns. Now, in a classic RPG, one encounter is a fight. In this RPG, however, one encounter is either coming up with an idea, or realising the idea in concrete form. Thus, we can’t use the classic RPG adventure structure, of fighting your way through the underlings until you hit the big boss.

Instead, I’m going to suggest that the equivalent of an adventure is a single creative project. Of course, you could do a project in two encounters: one to come up with the idea, and one to realise it. However, that wouldn’t be interesting enough, or long enough. The number of encounters necessary will depend on how long an encounter takes, but between four and twelve strikes me as a reasonable range. So, how do we get that many?

This is where I want to use something that Jeff mentioned in a comment on an earlier post: problem decomposition. Each project should be split into a number of sub-projects. These might be obvious “parts”: for example, the beginning, middle, and end of a novel. They could, however, also be things that can’t be separated out so easily. Thus, a novel might be separated into its protagonist, antagonist, setting, and plot. In order to create the novel, you have to create all the bits. The overall goal of the adventure should be a product with a quality over a certain standard, which will then achieve a goal set at the beginning of the adventure. The description of the adventure is, obviously, an important part of making the players care about achieving this, just as it is in any other RPG.

Another lesson that we should learn from classic RPG adventures is that railroading is bad design. This is the application to adventure design of one of our fundamental rules: the players and characters should make decisions that have an impact on the course of the game. Thus, the players and characters should be able to choose how to approach the central problem of the adventure, and it shouldn’t just be a choice between doing it the way the designer wanted, and failing.

Problem decomposition gives us a way to address this. First, once a problem has been split into sub-problems, those sub-problems can be tackled in various orders. There will, most likely, be some orders you can’t choose; if one of the sub-problems is “integrate all your creations”, for example, that really has to come last. However, complete freedom is not a necessary feature of an adventure, and not, in fact, a positive one; the paradox of choice, where having too many options paralyses people, comes into play. As long as there are multiple sensible options, even just two or three, the players are in control.

The order in which sub-problems are tackled needs to have an impact on the final result, probably by having the results of one sub-creation affect future encounters. The most obvious way to do this is to have it make the future creations either easier or harder. In addition, it’s probably best if each choice of sub-project makes some later projects easier and others harder, so that there isn’t an obvious best choice. If one choice is obviously better than the others, it’s arguable that there’s no real choice at all.

One potential problem to look out for here is the risk of cycles of improvement, where you do something that makes another step easier, then do that step, then use the benefit from that step to improve the first step, repeating until both are infinitely good. The rules must be set up to block this, but you also need an in-game justification. Fortunately, that is fairly easy to provide. Going back to redo the first sub-problem would normally require redoing the whole project, since you would lose things that you had depended on in the later creations. Rules to block these sorts of cycle should, therefore, not break suspension of disbelief.

There is a second way that problem decomposition can help us to avoid railroading. It might be possible to split an overall project into sub-projects in different ways. This needs to have an effect on the final project, in order to be meaningful, but as long as the final quality of the project is complex, this is easy to manage. Different decompositions can make different aspects of the final project easy or difficult to improve.

How much decomposition are we talking about? Let’s say that a project needs an idea for the whole project, and then integrating to produce a concrete result at the end. That’s two encounters. To have four to twelve encounters for the whole adventure, we’re looking at one to five sub-sections. One doesn’t make any sense, but we could have a very simple case where the idea for the whole project is trivial, as is integration, and the project itself splits into two parts, each of which requires two encounters. Otherwise, three to five subsections sounds reasonable. If you have three, A, B, and C, doing A first could make B harder and C easier, doing B first could make A easier and C harder, while doing C first could make A harder and B easier. There’s no obviously superior choice here, so that works for avoiding railroading.

If multiple decompositions are possible, we could also have decompositions with different numbers of elements. This might affect the time necessary to complete the project, which would be important if there was a time limit on the whole scenario. In this case, the method that takes more time could be easier overall. It might work better, if you can just get it all finished by the deadline.

It should be obvious from this discussion that writing adventures for this game is going to be fairly hard work, but that’s no different from writing them for any other game. One thing that is a little surprising is that there is still no sign that a gamemaster is going to be necessary. There is no information that needs to be hidden from the players to make the adventure interesting. Indeed, the characters should also know most of the game information in advance. In a campaign, there might be unintended consequences of certain decisions, and it might be better for the players not to know about those when making the decisions in question, but even that can be got around. The characters might not know, and it might be impossible to solve one problem without creating the new one. In that case, the players will have no choice but to go ahead, and they can role-play being surprised.

Finally, I said that there was a reason why the classic model was classic. What is that reason, and does it still apply to this adventure model? The reason is that the classic model breaks an adventure down into units. When designing an adventure, you can treat each encounter as a black box, with inputs and outputs, and connect those boxes together. Then you can design each encounter without having to worry about the details of the other encounters. Of course, you can add details that link the various encounters if you want, but that’s optional, and you only need to worry about it to the extent you want to. In short, the classic model keeps the complexity of adventure design under control, while allowing the experience to feel complex. That is preserved in this adventure model. The individual creations can be treated as black boxes, as only their qualities are relevant to the way that the later sections, and the whole adventure, turn out.

I need to actually write an adventure to work this model out in detail, but before I do that I want to look at character creation. I’ll do that in the next post.

Executive Abilities

The abilities needed to execute an idea are very likely to be different from those needed to come up with it (a possible exception is knowledge of genres), but what about the talents? Do they have to be different as well? I think that we can have some overlap, which is a good thing; characters should not have too many talents. All characters will have all of them, so if there are dozens, the character sheet will get a little too crowded.

So far, Empathy and Analysis have both been used once, while Synthesis and Creativity have both been used twice. We have another six “slots” in these rules, so it would actually be possible to balance perfectly, at three uses each, but that might not make any sense. However, it would be a good idea to try to spread things out a bit. We want to avoid having a must-have stat, or a dump stat.

First, Empathy makes a lot of sense as the talent that goes with improving Accessibility. The better you understand how people think, the better able you will be to make your writing accessible to them. You will know what they won’t understand on first reading, and know how to rephrase it to make it easier to grasp.

Embedding would really have to be Synthesis or Creativity, if we used the existing talents. Synthesis makes sense; you are looking for patterns in your idea and the cultural elements you are drawing on.

What about Technique? I think I’d like to introduce another talent here: Concentration, the ability to focus on something, and to keep all the aspects of it in mind at once. This makes sense as the talent governing Technique, because if you slip up once, the technical aspects of your work will suffer.

So far, Synthesis has three uses, Empathy and Creativity have two each, and Analysis and Concentration have one each. I probably want to avoid using Synthesis in the rules for reducing the resistances of the execution, but we are not terribly unbalanced at the moment.

Before looking at reducing the resistances, however, we should think about the abilities for improving the qualities. We really need at least one ability that reflects the ability to compose prose, and that ability should affect Technique. However, since there need to be two abilities that can affect Technique, maybe there should be two abilities, each concerning a different aspect of prose composition. Actually, two sets may be best, because of the candidates that occur to me.

One of the two abilities must also be able to affect Accessibility. One aspect of accessibility is the way that the work is written, the word choice and sentence structure. This would naturally be affected by the ability to compose prose in a particular language. We could call this [Language] Prose Composition, substituting the name of the language.

The other ability must also affect Embedding, so a Prose Composition ability specific to the genre you are writing in makes sense. Part of knowing how to write a particular genre is knowing the conventions and references to make, and this is not necessarily the same as knowledge of the genre as such. If you know about the genre, you might be able to say who wrote something, but not necessarily reproduce it.

However, knowledge of the genre, or, more broadly, of the culture, is a good candidate for the third ability. It affects Embedding, obviously, but also Accessibility. If you are aiming at people within a given culture, it helps to know what sort of things they know, and thus what assumptions you can make. The precise ability used here is rather open: it could be knowledge of the genre you are writing in, or knowledge of another genre, or more general knowledge of a culture. However, you shouldn’t be allowed to mix and match, because that would just create a confusing book. Once you have chosen a context for a novel, you have to stick with it.

This choice of abilities means that someone with good [Genre] Prose Composition and [Genre] Knowledge can write good works in that genre in any language she can speak. On the other hand, someone with good [Language] Prose Composition can write good works in that language in any genre she knows. This is, I think, basically reasonable. I’ve obviously concentrated on English Prose Composition, because I can’t write stuff in Japanese…

So, what about reducing the resistances? [Language] Prose Composition has to affect Embedding, and [Genre] Prose Composition, Accessibility, while [Genre] Knowledge affects Technique.

The last one is very easy to justify: Analysis and [Genre] Knowledge lets you see where the structure of the work is letting you down, by comparing to other works that you know, and the problems that they have. This also gives Analysis two uses. Before assigning the second ability, however, I’d like to look at the other two statistics. That might affect which ability looks sensible here.

I think that Concentration and [Language] Prose Composition can be justified for Embedding. Concentrating on how the words are put together might reveal places where you can put in more references without it becoming awkward, or places where you are wasting words, and could rewrite them to draw in more references.

So, the last problem is how to link [Genre] Prose Composition to Accessibility. Looking at the balance, we should use Empathy or Creativity as the talent. Empathy is already used for improving Accessibility, so Creativity is a better choice. So, what’s going on? Maybe you’re finding a new way to write something that makes it easier to include explanation or introduction, or that takes the emphasis off technical points and moves it back towards the characters. This would open up space for improving Accessibility again.

So, looking at the choices for second ability, I think [Language] Prose Composition is best for Technique. In many ways, it’s the purest technical ability here, so it should help you to find the problems in what you’ve written. That means that Embedding goes with [Genre] Prose Composition, which can be defended in the same way as [Language] Prose Composition, while Accessibility goes with [Genre] Knowledge. Your knowledge of the genre allows you to see a way you could make it easier for people to get into the book.

I think this makes enough sense for an initial outline. The talents all have a similar level of usefulness, as do the abilities, although [Genre] Knowledge is noticeably more useful than the others. What’s more, we still don’t need a GM. Might that actually be sustainable? I’m going to have to think about that carefully as I continue the design.

So far, I’ve outlined the “combat system” of the game; the system central to resolving the main individual challenges. However, I still need to think about the game as a whole. Just creating one thing after another would be as boring as just fighting one monster after another. How do I tie the challenges into a story?

Making it Real

As anyone who has ever created something knows, getting the idea, while essential, is not the hardest part, or at least not the part that requires the most work. Turning an idea into an actual creation is a major undertaking, and something that should also be central to this game. Actually, in some cases there is a further stage: music and dance needs to be performed, while industrial prototypes need to be put into production. However, this isn’t always the case. While a novel, for example, does need to be distributed, the distribution stage is entirely mechanical, and not really an interesting topic for the game. (Marketing is a different matter, and one that, again, applies to everything.) However, in all cases there is a step of turning the idea into something concrete, so I need to provide rules for that. Again, I’ll talk about art works to make things concrete, but the same basic structure should apply to scientific theories or inventions.

These rules should determine the quality of the final product, because when the activity they represent is finished, so is the product. The quality of the initial idea should influence the quality of the final product, and should do so quite strongly. However, the quality of execution also affects the final quality. Both aspects influence the final quality in all cases. A brilliant idea, poorly executed, is generally not a good work of art, although it does have redeeming features, and so is not as bad as a poor idea, poorly executed. This means that both aspects of quality must be reflected in the final result. For example, using the quality of the idea to limit the quality of the product will not quite work. The idea would not redeem a poorly-executed work.

However, there is a structure that could work for this. If we convert the qualities into dice, the execution could provide a dice pool, while the idea provides the number of dice you keep. This feels right to me; poor execution of a brilliant idea tends to have a lot more going for it than brilliant execution of a poor idea. It also allows works of art to be a bit unpredictable in how they affect people, because that would be determined by rolling the dice.

We could just use the quality as the number of dice, but that’s likely to produce an excessively large dice pool. If we divide it, however, we create thresholds in the difficulty range. Suppose that we divide by three, rounding up. In that case, a quality of 6 is no better than a quality of 4, as both give you two dice. The question is, is this a problem?

I think it may not be. If you have a statistic on 6, then it is probably worth trying to boost it, even if your chances of getting it up by 1 are very small. On the other hand, if the statistic is on 7, it might be sensible to give up. We can make this difference accessible to the characters, by saying that they can tell when something is very nearly working, so the decision can be in character. This will add a bit of tension, when characters try to gather as many bonuses as they can to get just one more point of quality, before the resistance of the idea increases and makes it truly impossible.

Another advantage of doing things this way is that it allows us to treat the execution of the idea as mechanically independent of the quality of the idea, which allows us to simply reproduce the basic structure of the mechanics for coming up with an idea. This is a good thing; as I’ve said before, using the same mechanical structure repeatedly tends to make a game easier to play.

There is one point of difference. I have to find a way to incorporate the Transparency of an idea into the process of execution. It’s supposed to be easier for a higher Transparency, so I can’t use the Transparency as a difficulty or resistance. I can, however, use it as a pool of points that can be subtracted from the resistance of various statistics during the execution process. The players can only use the points once, but they can split them between as many different cases as they like, or not use them at all on some rolls. They should probably only be allowed to reduce resistances, not the difficulties resulting from high qualities, but even so a high Transparency would allow you to boost the statistics of the execution quite substantially.

So, what are those statistics going to be?

A first, obvious, candidate is Accessibility, which measures how easy it is to get into the artwork. The Harry Potter books are very accessible, James Joyce’s Ulysses is not. Most people would say that the Harry Potter books are better, because high accessibility increases the number of people who can appreciate your work. However, that doesn’t mean that the Harry Potter books are, in fact, superior. (Having read both, I’d like to reserve judgement; they’re not exactly easy to compare. Yes, that does mean that I don’t think that Ulysses is obviously better.)

Another possibility is Embedding. This would measure how much the work refers to and draws on a wider culture. The more you do this, the more you can draw out the idea, but the less accessible the work tends to become. If someone has to understand the references to understand the work, the size of the potential audience drops substantially. If you do it well, however, you can do it without sacrificing accessibility, and make the work richer.

Those two statistics both refer to the content of the work. Let’s call the last one Technique, referring to how well the work is executed on a purely technical level. For a novel, this covers grammar, pacing, and characterisation, among other things.

If we go with this proposal, a completed work of art has five statistics: Originality and Resonance, for the idea, and Accessibility, Embedding, and Technique, for the execution. The Transparency is no longer important once the idea has been executed.

How would they work together for, let’s say, a novel?

Accessibility + Resonance could determine how appealing the book is on first reading, and thus how well it sells to start with. Accessibility + Originality determines whether people quickly think it’s a new thing, or whether it is shelved as derivative. On the other hand, Embedding + Resonance would be a good way to determine how engaging the book is on repeated readings, because the cultural embedding gives the emotional hooks more depth. The embedding allows the readers to discover more about the book every time they read it. Embedding + Originality could be used to determine how original the book seems on considered reading, against the whole cultural backdrop. Technique, on the other hand, might be used when the book is trying to defend itself against criticism. If your technique is good, it is harder for a hostile critic to tear the book apart.

This little exercise makes me think that the best way to come up with the statistics for a creation is going to be to look at how it is going to be used in the wider game, and then provide the statistics necessary for that. You may have noticed that some of the pairings above were a bit strained, but in a game that’s all about establishing a literary reputation in a hostile circle (say, a game set at the court of Heian Japan), the role of Technique might be very important.

In any case, I think they are good enough to serve as a worked example. The next step is to assign talents and abilities to the execution process, and that will be the task of the next post.

Helping Creation

The final part of my outline rules for coming up with ideas concerns ways that players can get more dice to roll. We’ve already covered some of these: I’ve said that good descriptions of actions will get additional dice. However, that’s a reward for player actions. I think we should also have character actions that improve the chances of success in the same way.

The immediate question, then, is what sorts of actions these should be. I think it might be best for them to be specific actions. That is, a character needs to have a particular item, resource, skill, or something in order to be able to take the action at all. Then, the character rolls a particular dice pool, keeping a particular ability, to see how effectively the thing is applied. The number of bonus dice you got would depend on the result of the roll, but the difficulty for a certain number of bonus dice would vary depending on the thing in question.

So, for example, an item might let you roll Analysis, keeping the character’s knowledge of a genre, and grant 1 bonus die if you succeed against a difficulty of 3, 2 if you succeed against a difficulty of 6, and 3 if you succeed against a difficulty of 9. A particular skill might let you roll Creativity, keeping the character’s knowledge of a topic, and grant 3 bonus dice if you succeed against a difficulty of 8, but nothing otherwise.

In general, these actions shouldn’t be able to help with all the actions involved in coming up with an idea, although some might. For example, a library of books from a genre might only be able to help with rolls using the character’s knowledge of the genre, whereas a meditation technique might only help with rolls based on Creativity.

It would be a problem if characters tried to use these actions every single time. This makes perfectly good sense for the characters, but it would mean that the players were taking the same actions repeatedly, and that’s a recipe for boredom. We could guard against that with simple “only use this once” rules, but that doesn’t fit with the other rules we’ve been using so far.

Instead, let’s give each action a resistance, which adds to the difficulties required to get a certain number of bonus dice. First, if you try to use an action, and fail to get any bonus, the resistance increases by the amount by which you failed. This ensures that, once it gets difficult to use an action, it will quickly become impossible.

Second, if you succeed, the resistance increases by the amount by which you beat the difficulty for the bonus you received. So, if, for the first example action, you rolled an 8, you would add 2 to the resistance, because you beat the difficulty for getting 2 bonus dice by 2. This means that successes will gradually make the action harder to use.

Bonus dice for the pool hit diminishing returns fairly quickly. For example, using Troll, we can see that if you are rolling 6 dice and keeping 3, your average roll is 14.3. That’s an improvement in average result of almost 4 over rolling 3 dice. On the other hand, at 9 dice, the average is 15.8, which is only 1.5 higher. The second three extra dice in the pool have less than half the effect of the first three. At 12 dice, the average is 16.6, about half the gain from the second extra three. This means that you don’t want to put all your bonuses on one roll; you want to spread them out a bit.

If an action becomes harder to use every time you use it, and you have a variety available, you are going to want to use as many different actions as possible. You probably would use at least one on every roll, but you’d try to spread them out, to get as much benefit as possible. Of course, there might be some critical rolls at which you would want to throw everything you had, but those would be special cases. In general, the characters would take different actions to prepare for each attempt to improve the idea, which is exactly the result we are looking for.

We might also want to allow for items or qualities that always add to the pool for certain sorts of action. As long as these are limited, they are interesting, and because they don’t need to be rolled for, they don’t create the boredom problem. An example might be a character who always gets 1 bonus die when using Creativity to create a work themed around the night. It might even be OK to do this for a whole genre, as long as the game was going to work on more than one genre. A character who gets 3 bonus dice to Creativity when working on a science fiction project is very likely to write a lot of science fiction, and to produce inferior work if she tries to enter a different field. That could well be an interesting piece of characterisation, as long as the game wasn’t just about writing science fiction.

So, we have three main ways to get bonus dice.

The first is player actions, primarily descriptions of the character’s actions. The better these are, the more dice the player gets.

The second is character actions. These actions could be taken by characters other than the one who gets the bonus, and thus they help to encourage cooperation between characters and players.

Finally, there are items and qualities that always add a bonus in a limited field.

Obviously, character actions are more limited than a constant bonus, so it should be easier to get access to such actions than to constant bonuses. Bonuses from actions can also get bigger than constant bonuses without posing too much of a threat to game balance, as long as the difficulty for the large bonuses is set fairly high.

The big advantage to this, from my perspective, is that it gives us many more ways to differentiate characters, and to have items affect in-game actions in a concrete way. This is going to produce this game’s equipment list, and list of character advantages and disadvantages, and give them a nice place to sit in the rule system.

At this point, I don’t actually want to go into more detail about them. The structure is clear, and that’s my current concern. Thus, this completes the first discussion of the creation of ideas. Next, I want to talk about realising the idea: turning the concept into an actual work of art.

Removing Obstacles

My general mechanic for coming up with an idea includes rolls to reduce the difficulty of a statistic or to increase the number of dice available to roll. In this post, I’d like to look at what you should roll to reduce the difficulty, and how you should role-play it.

First, a quick preliminary. I’m going to try out “resistance” as the name of the difficulty of one of the statistics of an idea, and save “difficulty” for the number against which you are rolling. Thus, each statistic has a quality and a resistance, and when you are trying to improve a statistic, the difficulty is the sum of the quality and the resistance.

So, the first thing I want to look at is actions that aim to reduce the resistance of a statistic. What sort of actions could they be? It would help if I clarify what, in the game world, the resistance is supposed to represent.

I think the resistance of a statistic represents the ways in which the current state of the idea restricts its future development. In developing the idea, the characters have decided not to pursue certain possibilities, and have also directed their thinking along certain channels. This makes it harder to improve the idea. So, when they reduce the resistance, they are breaking some of these restrictions.

Of the talents we have specified so far, Empathy does not look relevant. Creativity, on the other hand, does. It’s about thinking outside the box, and that includes thinking outside boxes you have made for yourself. What about Synthesis? That also looks relevant; it’s about putting together patterns, and that could certainly reveal a way to move forward.

If we tie each talent to a different statistic, then Creativity goes best with Transparency, I think. It’s about coming up with new ways to get something down in concrete form, so even though it doesn’t actually produce a way to realise the idea, it can suggest new ways to approach the problem, and thus reduce the resistance. Similarly, we could tie Synthesis to Resonance. Seeing the patterns in the elements that give the idea its emotional impact makes it easier to see how to increase that impact, because you can see how to add elements that complement what is already there.

We still need a talent for Originality, and I’m going to suggest Analysis, the opposite of Synthesis. Analysis is about breaking things down into their component parts, so here it would make clear exactly what was, and wasn’t, original about the idea, and thus show you where to make changes if you wanted to make it more original.

That gives us, once again, three talents for the three statistics, and, so far, four talents per character. However, I’m getting a bit uncomfortable about the way talents are tied to particular statistics. They might be better as different approaches to the same problem, but then, in order for them not to be functionally identical, something in the game statistics would have to make a difference to the way the talents worked. I’m not sure that this is a good place to increase the complexity, as it would probably mean more statistics for each idea. Thus, I’m not going to change this policy yet, but I still might.

If we leave the talents for now, the next question is about the abilities to use.

There is an obvious option that we can’t use. We can’t say that characters should use knowledge of a related genre that they haven’t yet applied. This makes sense; it would give a new perspective on the problem, and thus could help to clear away roadblocks. However, it makes the mechanics strongly path dependent. You would have to keep notes on which genres had been applied, adding to the record-keeping. I don’t want to make things more complicated than they are already, so this idea is no good.

What about keeping the same abilities, but switching the statistics to which they apply? That is, knowledge of the genre you are working in would apply to Originality and something else, knowledge of the topic of the artwork to Transparency and something else, and knowledge of other genres to Resonance and something else.

This might work. Knowledge of the genre is going to be very helpful in working out which parts of your idea are actually original, and thus it fits the way we’ve described reducing the resistance of Originality. It would seem sensible for knowledge of other genres to also apply to Originality, for the same reason, and knowledge of the topic doesn’t seem sensible this way round. It might help you think of something new to say, but it doesn’t help you work out whether someone has said it before in a particular artistic genre.

Again, given the way we have described reducing the resistance of Transparency, knowledge of the topic should be helpful. It will help you to think of other aspects of the topic that can be used to express it, and that might make the idea easier to express. The other ability has to be knowledge of the genre you are working in, because knowledge of other genres is already committed in two places, and, again, this makes sense. You can think of ways that other creators have approached a similar problem, and then go on to see whether you can apply that to your work.

Finally, what about knowledge of other genres and Resonance? Fortunately, this is also easily justifiable. You are trying to see the overall way in which the idea is inspiring emotions, and that will be assisted by being able to look at examples from other genres. Likewise, knowledge of the topic can help in the same way.

So, to reduce the resistance of Originality, you can use knowledge of the genre or knowledge of other genres. To reduce the resistance of Transparency, you can used knowledge of the topic, or knowledge of the genre you are working in. Finally, to reduce the resistance of Resonance, you can use knowledge of other genres or knowledge of the topic.

The fact that this is structurally the same as the rules for creating ideas is good, because that should make it easier to remember the mechanics. The fact that the abilities apply to different statistics may be confusing, but on the other hand having them apply to the same ones fails to emphasise the difference in the actions. The use of different talents might be enough to do that, so I’d have to consider applying the same abilities to the same statistics in both situations, but, again, that’s something to look at after playtest.

As for the description of the actions, we can use the same basic techniques as for the actions used to improve the ideas. For example, “I compare the idea to the concepts of several existing novels, teasing out the bits that can be said to be original.”. As before, good descriptions here would be worth bonus dice, from 1 to 3.

I think that gives me the basis for reducing resistance. Next, I have to look at actions that give more dice to roll.