Overcoming Resistance

As I mentioned in the last post, a character’s resistance will be a number tied to a description. To overcome it, personae roll dice.

The easiest way to do this is to have the resistance be the difficulty for the die roll, and, as yet, I don’t see any reason to make things more complicated. That still leaves the question of what the persona’s player should roll, and what she should keep.

I think that the player will roll the persona’s social skill. This will be something like persuasion or etiquette; the skill list will be worked out a little bit later. This means that someone with a high social skill is more likely to get a good result, but that charm and persuasiveness can only get you so far. Other things, the dice kept, are more fundamental.

The dice kept will be set by the strength of the persona’s relationship to the character, as long as the relationship is backed up by appropriate attitudes. Trust and love will be the most common, depending on whether the persona is trying to convince the character that something is a good idea for the character, or to do it because the persona wants it done. If the persona makes promises or threats, however, hope or fear might be appropriate. It isn’t really possible to use hate or doubt to get someone to do what you want.

Sometimes, this will be obviously hopeless. If the character’s resistance is 13 and the personae only have a two die relationship with her, they have no chance of overcoming the resistance. First, they need to either strengthen their relationship with her, or reduce her resistance. This is a good thing, because it reinforces the themes of the game. The personae need to build a relationship to get what they want, not kill things.

In many cases, however, rolling the dice will make sense, because the personae might succeed. What will be the effect of this roll? Right at the beginning of this blog, I talked about the effects that success on a die roll could have, and the need to avoid simple failure. In this case, I think there are two main choices: success might create an option, or change the context.

Success creates an option when it gives the personae another route towards the main goal, one that involves the cooperation of this character. For this to be appealing, the new route needs to be easier than the ones available beforehand. For example, if the personae need a ceremonial mask for a matsuri, and the character has one but is reluctant to lend it, overcoming her resistance lets them use her mask. If they cannot overcome it, the personae will have to make a mask, which will take time and effort.

Changing the context, on the other hand, will normally mean bringing more people within the benefits of reaching the goal. If a matsuri grants bonus dice to everyone present, convincing a character to attend means that she also gets the bonus dice. Of course, since characters do not actually use bonus dice, this would, instead, contribute to strengthening her relationship to the jinja, which in turn contributes to building the jinja up.

The players need to describe what the personae say to the character, and how the character replies. These decisions will be based on the elements describing the character’s personality, and on what the players have decided to have the personae do. This part of the interaction with the character is brief and simple, so the rules do not need to provide much structure. On the other hand, building a relationship is much more complex, as we saw earlier, and there, the rules do provide more structure for developing the character’s personality. What about reducing the resistance? That should also tell us more about the character, and so I will discuss it in the next post.

Character Resistance

The ways in which characters become obstacles to the personae will be defined with numbers, to fit with the rest of the mechanics. However, I think they will be defined with simple numbers, not dice.

This is because the active force in the game should be the personae, not the characters. The personae drive things forward, and make the decisions that change things. In mechanics, that means that the personae roll dice and succeed or fail. The success or failure of character actions depends on how well the personae do at preparing for those actions, not on a die roll made for the character. In any case, since there is no GM, there is no-one to roll on the characters’ behalf.

The first class of statistic is nothing but a number, representing the character’s resistance to a certain sort of action. For example, “Hates Masao” would serve as the resistance if the character were asked to do anything to help Masao, or work with Masao on a project. Similarly, “Blames the kami for his mother’s death” would be a resistance to getting involved with the jinja. The description of the resistance guides the players when they describe how the character reacts as the personae interact with her, and face her resistance. Personae can either try to overcome those resistances in particular cases, or reduce them in the long term. Both of these options will be considered later.

The second class of statistic is more complex. If characters never do anything at all without direct prompting from the personae, the game will not be very plausible, and players will find it hard to suspend their disbelief. Therefore, characters need to act.

In line with making the personae the centre of the story, however, even these actions will be reactions to things that the personae do. A character will be tied to a goal as a possible obstacle, and will take action that causes problems for the personae when they reach a certain stage in achieving the goal. The players, of course, all know that this will happen, because they have all read the scenario. The personae, therefore, also have a general idea that this character is opposed to the goal, but they do not know how to stop her. Investigating the character uncovers the reasons for her opposition, using the normal investigation rules, and the detailed reasons are created by the players as part of this process. The personae can then, if they wish, do things to postpone her interference, or to make it less of a problem when it happens. They might also fail at this, making her intervention even more of a problem than it would have been otherwise.

The game mechanics of the character’s opposition will be defined by the scenario, and may be modified by the actions of the personae. The description of her opposition, however, will be defined in play, based on the way the character and the situation have been developed. The scenario will provide elements that help to guide this definition, but they may not all apply, as the players may have taken things in a direction the scenario writer did not anticipate. Reacting to the character’s opposition may involve creating something, investigating something, or overcoming the character’s resistance, just as in the first sort of obstacle.

How, then, do personae overcome a character’s resistance? That is the topic for the next post.

GM-Free Scenarios

Why do characters come into conflict with personae? Because the characters want things that differ from what the personae want, or because the personae want things from the characters that the characters are not immediately willing to give. That, of course, is the in-game reason. The meta-game reason is that conflict with characters is one of the things that drives the story, and makes the game exciting. For the same reason, creation and discovery are difficult, and personae might fail, or at least have to accept something that does not quite live up to their hopes.

For creation, the players can decide what their personae are trying to create, and then use the rules to see whether they succeed. For discovery, the die rolls determine how useful the information is, and the available elements shape its content. For interaction, I think I can do something similar, which leads me to a conclusion about the fundamental structure of the game.

I am going to try to make it work without a GM.

However, my default position will be that the players are working from an existing scenario. This scenario might have been written by one of the players, or might be something that they have purchased. At least one of the players needs to be familiar with the rules of the game and the structure of the scenario, and things would probably go most smoothly if all the players were. However, in play, every player has a persona, and every player is working towards resolving the scenario in the way the personae want.

This means that a scenario needs to be playable by someone who has read it. For conventional RPG scenarios, this is not possible, because if you have read the scenario, you know everything that is going to happen. However, a lot of recent narrativist games do away with this assumption, and I’m going to follow in their footsteps. Kannagara will have scenarios that work better if everyone has read them, and reading a scenario should make you want to play it.

One way that Kannagara differs from most recent narrativist games, however, is in the level of rules. I don’t think it’s as complex as it initially looks, but it is a complex game. This means that the scenario will have to provide the elements necessary to support these complex rules. I suspect that, slightly counter-intuitively, this will actually make the game easier to play. Players will not have to decide what their personae do from an infinite menu of possibilities. Rather, at each point there will be a list of options to choose from. Players who are familiar and comfortable with the setting will, of course, be able to choose elements of their own, but that won’t be required. Creativity is hard, which is why I think it would be fun to play at succeeding, and why the game needs to make it easier to create in the game than to do it for real.

Of course, if the scenario is just a list of choices, people will not be inspired to use it. Motivation to play will come from the goals offered in the scenario. This might be possible transformations of the personae, discoveries about the world, or the creation of a great jinja. It could also be forming a good relationship with a kami, or resolving deep=seated conflict between characters.

In order for the game to be interesting, however, the scenario must also offer obstacles and conflicts. The rules for creation and discovery provide obstacles already; if a particular creation is defined as being difficult, then the personae need to find elements and improve their abilities before they can complete it. Characters are another source of conflict and obstacles. They might directly oppose the personae, but more often they have goals of their own that create an indirect obstacle. For example, the personae might want a character to do something she is unwilling to do, or a character might want to do something that will, incidentally, cause problems for the personae.

A character’s role in the scenario, then, is defined by what the personae can do for her, or how she gets in the personae’s way. That is not everything about the character, of course, but the rest can be defined by the players in the course of portraying her. The rest of the character can even have game mechanical effects, by defining what the personae need to do to win the character over. This is one way in which the scenario will be unpredictable, even to people who have read it.

While mechanics will be involved with the all aspects of the personality of characters, they are essential to describing them as obstacles, and that is the issue I will look at in the next post.

The Process of Intimacy

The rules described so far can be used for building relationships or changing attitudes. Although the game roles of these two are different, they are closely linked, and in the real world it would be hard to separate them. Therefore, the same kinds of actions can be used to modify either. So, the personae could use the rules to increase the extent to which a character loves or trusts them, or to build a relationship with that character.

Some parts of the process will, however, be different depending on the game mechanic involved. The most important difference is that a relationship is necessarily mutual, because of the way the game statistic is defined, while an attitude may be entirely one-sided. Indeed, it makes sense to allow actions that create a particular attitude in a group of people. It is certainly possible to build trust with a whole group, and similarly for awe. Love and hope may be a bit more difficult to imagine, but they are far from impossible. A whole group of people might have all four positive attitudes to a charismatic leader, for example.

If the persona has no relationship to the people who develop the attitudes, then the characters’ attitudes tell us nothing about what the persona thinks about them. Indeed, in some cases the persona might not even be aware that the characters exist as individuals. Again, think of the fans of a celebrity. The celebrity takes actions to encourage certain attitudes on the part of the fans, but while she knows that the fans exist as a group, she knows very few individuals among them.

Things are different if the persona is trying to build a relationship. In this case, the relationship is mutual, and the persona must be deeply involved with the character. Because a persona and a character have a single score to measure their relationship, the relationship score tells us nothing about their attitudes to each other. What is more, it seems strange to think that two people could have a very strong relationship, but no strong opinions about each other.

There is a simple way to address both of these points. We can say that a relationship cannot have a higher score than the character’s strongest attitude to the persona. This attitude could be a default attitude, if the relationship is only a weak one. Further, we will say that, by default, the persona has the same attitudes to the character as the character has to the persona. The strength probably does not matter, because attitudes will not be used to decide how personae act.

Normally, this is no problem. Personae will be trying to build mutual relationships of love and trust, because that’s the sort of game that Kannagara is. In some cases, however, things might be different. In particular, a persona might want to convince a character to trust her as a first step in getting him to stop lying all the time. In that case, trusting the character is a bad idea, but the persona needs to build a relationship in order to help. In such cases, a persona can try to build a relationship while having a different attitude.

This should be harder than trying to build a more straightforward relationship, and so should give the player a penalty to the number of dice she rolls or keeps. The penalty might be equal to the strength of the attitude, or maybe to the strength of the relationship. Either makes sense; the more strongly someone feels about the persona, the harder it is to avoid that feeling influencing you. On the other hand, as the relationship becomes strong, it is hard to maintain an asymmetry. This is something to be worked on in playtest, so I will make a final decision later.

Building relationships with characters is an important part of the game, but so is interacting with them more casually. As long as personae and characters do not disagree, no rules are needed for this: the characters just go along with the personae’s suggestions. When there is conflict, however, we need rules for resolving it. That is the next technical mechanical topic, but there is a broader topic I want to discuss first. I don’t think that Kannagara needs a GM.

Learning about People

What are the sorts of things that serve as elements when trying to find out about someone? The default assumption is that the personae are not spying on and manipulating the character, although that is possible, and should be supported by the rules at some point (although not necessarily in the core book). Instead, the personae are quite open about the fact that they are trying to build a relationship with the character.

The investigation elements will, therefore, include such things as “talk to her”. In fact, quite a lot of the elements will boil down to “talk to her”, so those elements need to be more closely defined. Recall that personae need to use several elements to build up an investigation, and can only use each element once. Each element, then, should not cover too much ground. “Talk about current events”, “talk about hobbies”, “talk about recent holidays”, “talk about religion”, and so on seem as though they would have a reasonable level of detail. Since the personae are being open about what they are trying to do, we should not neglect elements like “ask her what food she likes”. However, there are other useful approaches: “look at her outfit”, for example, because the clothes that someone wears tell you something about her. If the persona has been invited to the character’s home, “look at her bookshelves” is another option. Some elements here will be made available by the course of the game — by being invited into the character’s home, for example.

The other players describe the response. Here, the elements include things like “complains about politicians”, “plays tennis”, “visited Guam”, or “doesn’t know the difference between a jinja and a Buddhist temple”. Because the game is set in Japan, and most of the players will probably not be in Japan, the source material for the game will suggest some elements that could be used here, and the sorts of things that go together. Of course, the Japanese are also human, so there are a lot of similarities to people in other countries and players can also rely on their common sense.

The pieces of information that the persona puts together are things like “she likes cats”, “she is insecure about her appearance”, “she likes to think that she doesn’t care about food”, and so on. These are features of her personality, and while they can change over time, they are established at this point. These features do not determine what the personae should do to improve their relationship with her, but they do provide hints. They also help establish the character, for future interactions, and may also provide new goals for the personae. For example, they may decide to help a character who is insecure about her appearance to feel more confident about the way she looks.

Finally, the theory is, in this case, a description of the action the personae will take. This might be as simple as “buy her a cat”, or as complex as “hold a party with lots of delicious desserts, but where every dessert is from a different country, and captures part of the essence of the country, so that she can enjoy lots of desserts while telling herself that she is really investigating international culture”. (I think events like that are held in Japan quite frequently.) The players can generally just make these theories up, with the die roll simply determining how many dice they keep when they actually perform the action to boost the relationship. The game should provide some examples to give players hints and guidance, but this is another area that is within the realms of common sense.

One way I will incorporate Japanese culture into the game is by providing sets of elements that go together well for Japanese people, or that make good individual elements. For example, a lot of Japanese women want to own brand-name bags and clothes. This means that “wants to own brand-name shoes” is a good element for a Japanese woman. However, “has no interest in brand-name shoes” is also a good element, because it defines her as not following the general social trend. The important point is that a woman’s attitude to brand-name goods is significant, and, as far as I can tell, a man’s is not. Incorporating that sort of element into the game makes it feel as though you are playing in Japanese culture.

Before we finish the discussion of building relationships, I want to talk about the process. What happens while the personae are performing the actions that will create the relationship with the character?

Constructing a Theory

What do personae do with the information they have gathered? They put it together into a theory that tells them something about the subject they are investigating.

Putting the information together is another simple roll. The player rolls the relevant knowledge, and keeps the number of dice granted by the pieces of information. A persona can only incorporate pieces of information with a total incorporation cost of less than or equal to her score in the knowledge. However, the player can make this roll even if the persona has no score in the relevant knowledge; that just means rolling twice as many dice as she gets to keep, and keeping the bottom half. In this case, some of the pieces of information must have a negative incorporation cost, so that the total incorporation cost is zero or less.

A single persona may only try this once, for a given set of information. If she doesn’t like the result, she must gather more information before she can try again. It’s possible that the persona already knows more information than she can use at once; in that case, she can swap a piece of information out, so that she has a different set, and try again. Another persona may try with the same set, if she has the same score in the knowledge, and hope for a better roll. Because gathering information takes time, the normal situation will be for the personae to gather as much information as the best of them can use, and then each persona will try to make the best theory she can.

So, what does the result of the roll mean? In the context of building a relationship, the theory is an idea for an action to take to improve the relationship in question, and a good theory will provide more dice to keep. In other contexts, the theory may be very different. It seems sensible to build the theory out of elements, again, and have the information provide constraints on the elements available. Information that introduces complications, for example, will normally lead to theories with complications. More generally, information that a character likes strawberries will normally lead to the personae designing an action involving strawberries.

In contexts other than building relationships, these theories will often create options. A good roll indicates a theory that is a useful description of that aspect of the world, so a good roll for the theory could create one that allows a number of new activities. For example, there should be a theory that opens the option of transforming into a kannagi. A theory could also provide generally useful elements, which the personae can incorporate in their actions. The theory tells the personae that doing a certain sort of thing is effective, and so the personae can then do that, when it is appropriate.

An important side issue comes up here. The personae are gathering new information and putting it together into a theory; this is the sort of activity that increases your knowledge of a subject. Therefore, it makes sense for the persona to gain experience in the applicable knowledge from making these sorts of rolls. It is not actually possible to fail this roll, so the normal rule for gaining experience from using a skill does not apply. Instead, as a placeholder I’ll say that the persona gets 1 experience point for making the roll, and 2 if the total is more than 4 times the number of dice kept.

The rules here are heavily dependent on the elements, so in the next post I want to say a bit about the sorts of elements that would be involved in the context of building a relationship.

Interpreting

The next step in discovery is mechanically simple. The player rolls her persona’s investigation dice, and keeps the number of dice indicated by the elements of her investigation. The result of this roll is the pool of points that the player can use to describe the information that her character has discovered. Again, this is done in terms of elements.

Each element has a discovery cost, which is the number of points needed to make it part of the information discovered. Standard elements also have an incorporation cost and a number of dice, which are used when generating a theory in the next stage. A low incorporation cost and a high number of dice is good, because that will tend to produce a better theory. Naturally, a low incorporation cost and a high number of dice mean that the element has a high discovery cost.

Once an element has been discovered, it is part of the description of the subject. If, for example, the persona discovers that a character likes cats, then that character does like cats. The character may change her preferences over time, but even if she does it will at least be true that she did like cats. This means that some elements will exclude others; a character who likes cats cannot dislike them as well. Of course, a character can like cats and dislike dogs, so these restrictions will often be simple, and a matter of common sense.

When the personae are discovering something about a subject that is not a part of the real world, however, these links between elements become more important. If an element is a feature of a subject, it might make some impossible, and others more or less expensive. The players are, effectively, creating the topic as they investigate it, so these relationships will mean that the topics will fit into the game background. This, however, is a complicated issue that doesn’t apply to investigating people, and so I will come back to it later.

Some elements have a relatively low incorporation cost or high number of dice for their discovery cost, because they also introduce complications into the personae’s lives. For example, discovering that someone is a huge fan of first edition D&D books is useful, because it suggests an ideal gift. Quite an expensive ideal gift, however, and one that may not be easily available, so that adds story potential to the creation of the action that will build the relationship.

While the number of dice granted will always be positive, and normally not more than one or two, the incorporation cost can be negative. This is necessary because of the mechanics for building a theory, as I will discuss in the next post.

Not all the elements discovered in this step need to contribute to a theory. At least some of them should, because the theory is typically the purpose of the investigation, but there are other options. This is one of the places in which a persona can discover a special element that grants a bonus to some other task. These elements have discovery costs, but their other statistics may be quite different.

Any subject has a number of elements available for discovery. Only the ones that have been discovered are known to be true of the subject, however. Other elements may be true of different examples of that sort of thing (different kami, for example), and elements that are not inconsistent with what is known may also be true of this subject. It is possible that the personae have simply not learned everything about it yet. Nevertheless, those elements cannot be used in play because, even if they are true, the personae do not yet know that, and so cannot use them.

The final stage of discovery is creating the theory. How does that work?

Investigating

The investigation phase of discovery does not involve any die rolls. Instead, the players describe how the persona is investigating the subject, and describe something of what she finds. This is tied to the game mechanics by elements.

The elements of this phase describe things that the persona does to investigate. Each element has two statistics: a cost, and a number of dice. A player may only incorporate elements with a total cost equal to or less than her persona’s score in the investigation ability. That is, if she has a score of 4, the total cost of the elements used to describe the investigation must be 4 or less. In addition, because the elements describe how the persona investigates, each can only be used once. It isn’t a single action; it represents the application of a certain approach to the persona’s investigation of the subject.

The higher the number of dice, the better, because these are the dice that the player gets to keep when rolling for the next stage of the discovery, when the persona puts together what she has discovered to provide information that will be useful in developing a theory. However, the default investigation elements, available to any persona, will have a higher cost than the number of dice that they provide. The simplest benefit of a higher score in the investigation ability is that it allows a persona to use more of them, and thus have more dice to keep. However, a high score may also allow access to a limited number of more effective elements. Personae might also gain access to effective elements in play, as mentioned earlier. For example, if a character gave the personae her diary in play, the element “Read her diary”, which might cost 1 point and give 4 dice, would become available to all the personae. For general elements, personae might learn effective investigation techniques.

While the persona’s player describes how she is investigating, the other players describe what she finds. This is important, to keep all the players involved, and also because the investigation might involve the persona talking to characters. In such a case, the other players take on the roles of the characters. If the persona is investigating something that is not alive, the other players just describe what she finds. The players introduce elements when saying what the persona uncovers.

These elements do not interpret the findings, because that is the job of the next stage. They simply describe the immediate results of the investigation. For some things, these elements will be defined by the game, and there will be suggestions for everything, but for many topics the players can make them up. If the personae are trying to find out what a character likes, for example, the players can just say things that are plausible things for a person to say in those circumstances. They do need to define the element, however, because that will be important in the next stage, when the persona works out what she has discovered.

Every element must be played out, the players describing what the personae and characters do. The investigation ends when the persona’s player has described how she is applying every element, and the other players have described what she finds. At this point, there should be two or three elements on the table, giving a superficial description of the evidence that the persona has found. (I find that I’m tending towards thinking of the elements as cards, in which case they might literally be on the table.)

Next, the persona describes what she has discovered, but that is the next stage, and will have its own post.

The Basis of Discovery

Discovery is also supposed to be an important part of Kannagara. In the grand concept, the discovery in question is things like discovering how to transform yourself into a kannagi, but discovering what sort of food someone likes so that you can cook her the perfect dinner falls under the same general category, and that’s the case that has come up first.

This is also a somewhat easier case than the profound discoveries about the universe, or about the secrets of the setting, because it does not raise the problem of whether the players should know things that the personae do not. The preferences of a particular character can be decided while playing through the discovery without undermining the consistency of the whole setting, and the sudden realisation of the truth is not, in this case, supposed to be a big moment for the players. There’s also no problem, I think, with accepting that the success of the discovery, measured in terms of the number of dice the personae keep for the assessment, reflects how well the personae understand the character’s preferences, and the players will not feel that they “need” to have the same understanding. That is, the players do not have to describe the character’s preferences in great detail just because the results of the investigation mean that the personae understand all the details.

I do want to use the same basic rule structure for all kinds of discovery, so this is a good place to define the first draft of those rules. Obviously, they could change beyond recognition in later revisions, but I have to start somewhere.

The discovery rules have three stages. First, the player describes how the persona is investigating the problem, and describes something of what she finds. Other players may also contribute to the description of what the investigating persona finds, even though their personae are not involved in this part of the investigation. No dice are rolled at this stage. Instead, the elements incorporated into the investigation give the player dice to keep at the next step.

In the second step, the player rolls her persona’s investigation skill, and keeps a number of dice determined by the elements that were incorporated into the investigation description. The result of the roll determines what information the player can describe the persona as having discovered. This information also takes the form of elements, which are used in the final step.

This third step is the generation of a theory. Once all the personae who want to have gathered information, and shared it with each other, each persona can try to create a theory. The number of dice to keep is determined in the previous stage, while the number of dice rolled is the relevant knowledge. This theory is primarily described in terms of game effects, and a higher result on the roll results in more favourable effects. This represents a better understanding of the situation.

This structure keeps all the players involved as much as possible. Everyone is involved in the investigation stage, and every persona gets her own information stage. Finally, each persona can try to generate a theory, so everyone can be involved at this point, too. Personae with better knowledge and better investigation abilities will, however, do better at it than others, and have their chance to shine.

I will describe each of the stages individually in the following posts.

Putting Numbers on Bonds

In order to make the rules for building relationships definite, I need to start attaching numbers to it. The best place to start is with the strength of relationships, since everything will develop from there.

A strong relationship is, let us say, a score of 6. This allows for a lot of differentiation between relationships; the average result of keeping 2 dice is only one higher than the minimum possible result from keeping 6. Of course, relationships can be even stronger than that, but we will assume that 6 is a normal maximum, and probably the default value for a father’s relationship with his children.

When you are building a relationship, you can try new positive actions at any time. Even if they do not improve the relationship, they will not make things worse, and I want to retain that in the game. The only limit is time. For a rough approximation, suppose that you are keeping the same number of dice for both the action and the character’s reaction to it. The maximum possible roll is 6N, where N is the number of dice you are keeping, and the maximum possible reaction total is 12N, because it is doubled. In order to improve the total, the difficulty must be no higher than 6N-1. If you roll the maximum possible result on the reaction total, the difficulty for the next action roll is 6N-1 if the running total, R, is such that R-(12N-R)=6N-1, or 2R-12N=6N-1, or 2R=18N-1. So, the highest possible total is 9 times the number of dice you are keeping.

This is too hard to reach, however, so we should set things a bit lower. 8 times the number of dice would be reasonable. Let us say that you need to be keeping 6 dice to build a really strong relationship, and set the difficulty at 8 times the level of relationship at which you are aiming.

How would we get that many dice? First, there’s the question of creating the action. We could use the same difficulty scale as the matsuri, with a base of one, so that just doing something generic lets you keep one die, and adding personalisation lets you keep more. In that case, getting to keep six dice would have a conception cost of 13 and a difficulty of 32. That should be just about possible for people keeping three dice to devise the action, because they can get the difficulty down to 27, and it’s reasonable for people with four. That means that, according to the rules, people who are decent but not incredible can create the actions that let them build a really good relationship. This is reasonable; building good relationships is not the province of an elite few in real life.

What about making the action appropriate to the character? That brings us back to the question of investigation and discovery, and so needs to be the subject of a different post. However, we do know that getting six dice should be just possible for people keeping three dice, and a reasonable target for people keeping four, again because most people can build good relationships. That will help me to fix the numbers.