[Mayhem] Resolution System

[mogunus]

I am having a particular problem hashing out how my resolution system is going to work.

My resolution system should facilitate simulationist play. I would enjoy some non-trivial Exploration of System here. I am uninterested in integrating Drama, I would like a Fortune and Karma based system.

These are the aspects of a task or conflict which I should like to represent in the resolution system:

   1. Variance: How random is a task? Some tasks or conflicts, like chess, are very deterministic. Others are more subject to random effect, even for skilled practitioners: writing or painting are examples.

  2. Skill: Characters can and should be ranked according to their skill in a task. I am unsure whether I should make a distinction between "logarithmic skill increase" and "linear skill increase," and use one or just the other.

  3. Nature: characters have different physical and sensory limitations. The resolution system should take this into account. A hale and healthy character will resist disease longer or march harder than a sickly one.

  4. Difficulty: How hard is the task? This presents several problems. Are tasks to be rated on a absolute or relative scale? I lean towards absolute, because it seems like a relative scale would add an extra layer to the system which is possibly un-necessary, and it would be easier to assign and deal with absolute values for tasks (a level 2 door, a level 3 chess opponent).

My major concern is conflicts between in-game entities of vastly different scale, say, a Dragon and a Human. I see no way to represent the vast scale differences between these entities save drastically different numerical values for "Skill" and "Nature." This may or may not be a problem.

This is what I have come up with so far:

Each character has physical and mental limits. They are rated on a scale from 1 to 20. The limits are strength, endurance, dexterity, fluid intelligence, crystallized intelligence, perception, and will. Each task has an associated limit.

To resolve a task with difficulty n, a player rolls x d20's, where x is either 3 (default, untrained value) or the character's skill rating in whatever they are attempting. Out of these dice, the player counts those which are equal to or less than the character's associated limit, and compares this count to n. If it is equal to or greater than n, the attempt succeeds.

There are a few serious problems with this system, the most glaring being that it does not scale, and breaks around the edges. A character with ability of 20 will always succeed, if they are rolling enough dice to do so, and a character with ability 1 will always fail. I could constrain ability ratings to a smaller range than the d20, or base the system on a larger die (a d30 perhaps?) but this feels awkward, and I have no concept of total "orders of magnitude" difference in the mental and physical limits of in-game entities. It would be nice to represent these limits such that only an incredibly exceptional human has a chance to draw or defeat a dragon in a game of chess (assuming that dragons have average fluid and crystallized intelligence far beyond that of a human). It seems that a 2 - 18 (constrained to remove the "always failing/always succeeding" problem) range of limits is not enough to represent these drastic differences.

Currently, I do not know how I should represent task variance.

Any suggestions, or pointers to systems that achieve any of the goals listed above, would be very appreciated.

[mogunus]

More on the objectives here:

I am interested in producing vivid descriptions of in-game events, which are embedded in a rich framework for making player choice matter. So, a blow-by-blow depiction of a sword-fight, where at each juncture the player has multiple choices which can influence the outcome, or a step-by-step account of a magic ritual, with varying in-game effects based on character actions.

I am not so very committed to the system above. Another option that I have been considering, more karma than fortune based, is to enumerate for each activity a "decision tree," as in the computer representations for chess or checkers, and take the player through selecting their next actions in this tree. Are there any games which use this format explicitly? Of course, fortune could be integrated by giving each branch a weight which represents probability. In conflict, nodes could be set up in a type-based defeat/victory hierarchy, with satellite data for consequences of node selection.

Note that I do not want to integrate hit points into the game. I am more interested in damaging physical and mental limits directly, as well as rolling on tables for specific wound effects as this happens. Combat should feel swift, brutal, and deadly, with lasting injuries.

Similarly, psychological conflict and fatigue will directly affect the character's mental limits. I do not view limits as "nature" ability scores which are mostly unchanging: rather, they are measurements of the character's current physical and mental limitations.

Any comments or suggestions very welcome.

--Marco

[markhaselb]

   1. Variance: How random is a task? Some tasks or conflicts, like chess, are very deterministic. Others are more subject to random effect, even for skilled practitioners: writing or painting are examples.

You can achieve this by using replacing D20 with 3D6 (or maybe 2D10) for deterministic resolutions. While on the D20 every number has an equal chance, the chance for a number of eyes in the area of 10 to appear rises the more smaller dice you use.

  2. Skill: Characters can and should be ranked according to their skill in a task. I am unsure whether I should make a distinction between "logarithmic skill increase" and "linear skill increase," and use one or just the other.

Using only one of the two options is better in my opinion. The more different systems you use, the more confusing it will be for players.

4. Difficulty: How hard is the task? This presents several problems. Are tasks to be rated on a absolute or relative scale? I lean towards absolute, because it seems like a relative scale would add an extra layer to the system which is possibly un-necessary, and it would be easier to assign and deal with absolute values for tasks (a level 2 door, a level 3 chess opponent).

I think this is a matter of taste, as long as you treat relative values as differences of values and not divisions.

To resolve a task with difficulty n, a player rolls x d20's, where x is either 3 (default, untrained value) or the character's skill rating in whatever they are attempting. Out of these dice, the player counts those which are equal to or less than the character's associated limit, and compares this count to n. If it is equal to or greater than n, the attempt succeeds.

I head to read this part a few times to understand it. As far as I understand it goes like this.
- roll X + 3 D20 ( why would you roll only one die if your skill is 1, but roll 3 dice if your skill is 0?)
- count the dice that show numbers below the corresponding ability for the task. Let's name this number S for successes.
- compare S to difficulty N, if S is a value of minimum N you succeed
Trying to roll achieve low die results and in the next phase of the resolution trying to achieve high results seems confusing to me. Maybe there is a way to use high results in both phases. It is not that big of a deal though.

Concerning the dragon vs. human chess problem: it does not seem like a problem to me. Make a contest, both parties roll their chess check, the one with more total dice successes wins. This will most likely be the dragon because high attributes generate more successes. Maybe you grant additional dice for high ability scores as well: one more die if the ability is greater than 25 or something like that.

Note that I do not want to integrate hit points into the game. I am more interested in damaging physical and mental limits directly, as well as rolling on tables for specific wound effects as this happens. Combat should feel swift, brutal, and deadly, with lasting injuries.

By directly damaging the abilities you have the solution for your other problem (how does low health affect skill checks?). When designing your combat system try to avoid to many skill checks with your above resolution that don't result in effects on the opponent. Some systems have high chances of missing opponents and this gets boring with rolling one die only. The more time you have to spend rolling dice, the less swift feels combat.

You want to put a lot of effort in detailed systems for combat and other thing, but does this reinforce the focus on aspects of the different cultures and myths in your game? It seems that you drift off from what you said in the Power19-thread and the society aspect of the game is reduced to a nice gimmick instead of being the main theme.

It is good to see new ideas though, I am looking forward to reading more

[mogunus]

Thanks for your comments.

You want to put a lot of effort in detailed systems for combat and other thing, but does this reinforce the focus on aspects of the different cultures and myths in your game? It seems that you drift off from what you said in the Power19-thread and the society aspect of the game is reduced to a nice gimmick instead of being the main theme.

This is something that I have been worrying about. My current reasoning goes: I have warrior, priest, sorcerer, farmer, hunter, and crafting roles. Each role should be supported by a nontrivial set of rules that allow players to do some exploration of system, and feel as if their choices matter and their in-game actions obey some set of consistent laws. I'm working on a detailed set of rules for farming, for example, derived from my prototype task resolution system. If I simulate all of these activities at about the same level of detail, then they serve to reinforce the sense of in-game causality and differences between player roles.

In my village scenario, a couple players are going to be warriors, one a farmer, and one a priest. The rules-heavy resolution system should serve to emphasize the differences in their training and activities, while providing an about equal depth of activity for each. More importantly, meals, festivals, ceremonies, and full-village events would be the focus of group player activity.

Alternately, I could see abandoning this sort of program, and using a more abstract resolution system, perhaps only the core of what I have above. While I want activities associated with player roles to feel "real," I do not want the activity simulation to detract from the game's focus on personal and communal myth. I am very willing to sacrifice details in activity simulation for this goal.

Also note that the role advancement trees have social/cultural aspects to each node, as well as activity simulation aspects. Idea being, it is hard to change what one is capable of without also changing how people view and treat you, inside and outside your profession.

With the explanation above, does my desire to provide detailed rules for player activity make sense? Or does it still seem like it would detract from the myth/cultural focus? I am quite unsure here. Depending on response, I'll either pick one, or "branch" the game and playtest both variants.

Further:

I really like varying the die type to reflect task variance, that's a very nice idea.

markhaselb's read of the resolution mechanic is correct.

I am trying to design a quick/brutal combat system, with a high hit chance per roll. I feel like at least some attention should be spent on this, because there is quite a lot of mythology dealing with violent heroes, and a warrior ethos, but I don't want it to detract from examining society's myths.

An example of the sort of Scenario I want to support: players are members of a hunting culture. Slowly, an agrarian society encroaches on their territory. A physical and cultural clash ensues, over several generations, and the peoples eventually form a hybrid culture. Over the course of the scenario, players may select characters from either "side," and play out each character's lifetime individually. Does this seem too deterministic? The point to simulating hunting and farming in detail would be to get an idea of the food supply available to each culture, so as to get an idea of their population and health over time. On the other hand, this could be unnecessary... Thoughts?

[Bill_White]

You might consider starting from situation generation: how frequently do dragons and raiders show up, how often do farmers face droughts and locusts, and so forth. Then think about the "standard task sequence" required to deal with it. And maybe it's not about skill rolls but about expenditures of some in-game currency that can be traded around: the warrior's prowess, the priest's blessing, the farmer's labor, and so forth. In other words, model the sociocultural economy game-mechanically and the rest will follow easily. This is a lesson it took me a long time to learn, but I give it to you for free ;-)