Persistent fantasy – system update

27 August, 2008 at 11:37 am (game design, persistent fantasy) (, , )

These are the clarified and collected rules of my default house system, still somewhat in progress.

Big picture

The design goals of this system are, in no particular order:

  • Sword and sorcerish fantasy adventure.
  • Meaningful one-shot games.
  • The setting and the characters have continuity, even though the players present in any one game may radically change.
  • Playing the game generates the setting, the characters and the rules.
  • Planning ahead is futile and impossible, or at least very hard.
  • Players have power to shape the fiction.
  • Character concepts are protected, in that breaking them is hard.
  • Conflicts should be detailed, not tactical.

Cycle of play

  1. Gather players. See how many participants there are. Take half this number, round down, and consult the list (my list is visible in the persistent fantasy page). Select the topmost characters whose players are present, though a given player can only get one character from the list to play.
  2. Consult a relevant random generator (my generator: http://random-generator.com/index.php?title=Fantasy_oracle_compilation). The players who are thus far without characters generate new ones implied or explicitly mentioned by the random entries. Should some named character not controlled by another player be eligible, playing that one is recommended, but not obligatory.
  3. Generate a situation based on the random elements and characters: What is happening, why are the characters involved, and what do they want?
  4. Play.
  5. Update the list and character records.
  6. If such is the habit, also update setting info.

Anatomy of characters

  • Starting player characters have three traits. At most one trait may have the value of six. All have integer values between 1 and 6. Values 1 and 2 indicate a minor trait, 3 and 4 a significant one, 5 a major one and 6 something potentially legendary.
  • Starting non-player characters have an arbitrary number of arbitrary traits. All must have positive integers as values.
  • All player characters must have a name.

Anatomy of traits

All traits work as defined below in the section that relates to resolution. Some have additional qualities. Examples follow. (Ksym, corruption is built for you, if you are interested.)

  • Aspirant to the throne: Increase this trait by one whenever the character does something that will prove useful in evidently grabbing the throne. Inciting a rebellion in the streets, blackmailing noblemen, gaining favour of the church, being of noble birth, deterring an invasion in the borders. It can be reduced by one as a consequence of failing at the listed tasks. Aspirant can be rolled when the character is trying to take over the throne, or at half value when negotiating with the nobles. After winning the game of thrones the trait is replaced by Queen/King of the realm or similar trait at the same value. Any trait that signifies a goal can be handled in similar way.
  • Corruption: Corruption is any source of power that takes over the character when used. A demon whispering advice, a blade with something evil bound to it, the power to make the dead walk. Using the corrupted trait means that its value increases by 1. (This does not apply to others using the trait, only the character who has it, though some corruption may spread to other characters who use it against the person wielding it.) When the value of the corrupted trait is greater than that of any other trait, it transforms the character into something else. Demonic whispers trait may turn into champion of chaos. Necromancy turns into lich.
  • Creeping doom: Any effect that slowly consumes the character. A disease, say. Increases by 1 every session. Trait may evidently transform into something else or just totally cripple the character. Curing it may be useful.
  • Lycanthropy: Any trait that can be drawn strength from, but without the condition becoming more severe. Use the trait and certain effects will come to be; your character is wounded in battle, you use the werewolf trait, win the combat, and later the character wakes up naked in a graveyard next to half a corpse of human, for example.

Names and the list

Some characters are named (including all player characters). Anyone can name a character in play. Some named characters are on the list.

While a character is on the list that particular character can’t be permanently removed from play. Should one be killed, it will later become apparent that this did not actually happen or that the character has returned to life or is now a demon who just happened to be summoned back to this world. Also: Undead.

Any character on the list is owned by some player (or GM). Unless a player explicitly says otherwise, other players are not to play that character. Any player can renounce a character; this means that any entries where that player and that character are linked are removed from the list. It is recommended for one to not renounce characters.

Named characters who are not on the list can be played by anyone, regardless of who previously played them. They can be killed permanently in game (though doing this without the character entering the list is hard) so that they actually will not come back.

Named characters not on the list should be possible result of the random generator. Named characters on the list may not be.

Unnamed characters have some traits and function as other characters do, but they do not enter the list. An unnamed character can be named at any time in play by any participant. This may be done as a reaction to the character failing a roll, so that the character immediately enters the list.

A brief respite

Some player may want to retire a character for a while, but still retain control of that character. The fiction must offer a suitable opportunity for doing this; for example, a character building a cabin in some forlorn woods and staying there. Or entering a large city incognito.

This can only be done if the character has at least one entry on the list. Remove all entries featuring the character from the list. After the session add the character once to the bottom of the list.

If a character is controlled by several players, all of them must agree to the respite. The process works as above, except that the character gets one entry per player to the bottom. The order of the entries is determined by their previous order; the player who previously was closest to the top with that character gets the first entry, and so on. Players can agree to different order.

I will not abandon you

Some characters may stick together, no matter what. A D&Dish adventuring party, a married couple, brothers in arms, a group of soldiers. Pets, cohorts, and such also qualify. The mechanical term for this will be “group”. Grouped characters must have traits linking them to the group, mentioning other members by name. Not all characters need to be thus linked, but any two characters must be linked through other chars. (A linked to B linked to C linked to A, for example.)

Grouped characters will only enter play if all the relevant players are present. If even one of the characters would enter play due to being on the list, then all will. The top-most name of every grouped character is removed from the list.

The actual procedure when using the list is to select characters as normal so long as a group is not encountered. When one is, check to see if all the players are present and if none of them are yet playing another character. If both of these conditions are true, then the relevant players are assigned the characters in the group. Names are removed from the list as above. The players hence selected count towards the total number of player assigned characters before play, and may exceed that number.

(Note to ksym and others playing: This replaces bound characters as a concept.)

Actually playing

By this point, every player has a character and the fundamental situation has been created. Characters are motivated or just plain involved.

Play happens in freeform manner; players tell what their characters do, participants tell what happens around them (should a game master exist, this is likely to be done by them). If a given player is otherwise uninvolved, playing and possibly naming an NPC is a good idea, if that particular player is up to the task.

Adjusting traits

Characters can gain and lose traits in play. This can be in the way of flashback (“When Moh was young he had to fish for his own food, so he has the trait fisher 3. Okay?”) or in-game events (“Since you killed those outlaws the word has been spreading. Take vigilante 2 as a trait.”), or mix thereof (The game starts at desert and my character has been wandering around for a while. I’ll take desert survival 1.)

The above applies to adjusting the values of traits, too. Extensively practice a skill, get better at it. Be in location where a particular skill is never used and it might go down (if someone bothers).

Resolution system

(I use six-sided dice. Other dice can be used, as long as there are enough dice to go around. Everyone must use dice with the same number of sides.)

Should a situation where both (1) the outcome is uncertain and (2) player character is in risk emerge in play, dice come to play. One can use the dice in other situations, too, but inconsequental rolls should be avoided. They break things.

For every involved character some player tells what the character does and names a useful trait of that character or a trait that other character has that can be exploited in this situation, and further that same player explains what the character does, including how the trait is used. For example: “I know this land and the best route through it (trait Born in the desert 4).” or “I meditate and consult the spirits, asking them to reveal the lay of the land (trait Shaman 3).” A clearly relevant trait gives its value in dice. A partially relevant trait gives half the value, rounded down.

Any number of traits can be introduced to a given conflict, but every time one must describe how the trait comes to play. Turns should be taken such that every player, in some order, adds more details to the conflict. This is the way through which play creates fiction.

A player can opt to remove a name from the list to get more dice. The name must be of the character in conflict and it must be linked to the player currently playing that character. Removing the top-most name on the list gives 3 dice, while the name on the very bottom gives 1 die. Every other name gives 2 dice. (Unless otherwise mentioned by the relevant player, the lowest name that gives 2 dice is removed.) This can only be done once per character per conflict. Description should be related to good luck or other unrelated factor.

If some characters have no dice, those character get a single die and the pools of everyone else are doubled.

The dice are rolled.

Interpreting dice

Some characters oppose each other, some do not. They are respectively called opponents and allies. (In this case, someone really is on your side or the other side; mechanically, there is no way of being neutral.) The relations need not be transitive, but they are symmetric. That is: Any two characters are either allies or opponents, but two characters allied to same char can still be enemies. A likes B, B likes C, A hates C (this is only relevant if there is someone for B to oppose, too).

The dice are rolled. If any opponents have dice showing the same number, every character in the conflict removes one such die from table. Continue as long as any opponents have dice showing same numbers. After this process is done, there will be one character or several allied characters who have the highest die still on table. This character, or these characters, are winners. They have a number of successes equal to the amount of dice they have that are higher than all the dice of their opponents. All characters that are not winners get their name on the list. (Order arbitrary, but generally the characters who fare the worst should be entered first in order to remain true to the game’s principles.)

In case of tie (no dice remain on the table),  there are no winners. The conflict must be started from beginning, but the situation has somewhat changed. Nobody gets to the list.

The participants playing the winning character, or winning characters, now have the power to describe what happens in the fiction. The participants playing the losing character(s) have veto power, but using it means they take harm equal to the number of successes of the one who described the events. In case of harm the situation stays unresolved (but it has changed regardless due to the way participants described their characters acting). Dice can be used again or the situation may dissolve.

Harm

Characters can only take harm when losing conflicts. Harm stacks. Player can’t invoke any traits with value equal to or less than the amount of harm that player’s relevant character has. For example: Character with 4 harm can only make use of traits with value 5+. This includes exploitable traits of other characters.

Any character with harm equal to or greater than that character’s highest trait is removed from play for this session. Maybe the character is unconscious, dead, lost, or just got bored and walked away, whatever is appropriate to the fiction. Characters can’t recover from harm in play. All harm is removed from all characters between sessions.

In play

In actual play conflicts usually have less than five participants. Mobs of people or swarms of creatures are better represented as single entities. (Peasant mob traits, an example: Torches and pitchforks 5, mob mentality 4, burn the witch 4. Peasant traits, an example: Peasant 3.)

The way dice work is actually pretty easy to show, but hard to explain. What is notable is that there are no tactical decisions to be made, except perhaps in fiction. What is also notable is that the dice practically never produce a tie; it would require that all the dice were removed from the table, which is extremely unlikely, but possible in few cases.

Also, in conflicts it is useful to check who is winning right now to provide inspiration for the narrative. I am not inclined to explain the process here, however, but suffice to say it does not make a difference on the level of crunch.

Players usually accept the offered narration. If it is unacceptable, negotiation is possible instead of taking harm. It is recommended to suggest what might happen to other participants. Don’t try writing a novel, just keep the game moving in some simple way. Embrace the obvious, because what is obvious to you might not be so to others.

It is possible to grant traits as a result of conflicts. It is usually a good way to harm other characters, for example. General guideline: The value of an added trait should be roughly equal to the amount of successes one got. Charming someone with 2 successes might give the target trait “Trusts [whoever happened to be the charmer] 2”.

Running and playing the game

This game does not actually need a gamemaster, though one is very useful for adding adversity in game. If all characters are on each others throats anyway, GM will play a smaller role. Co-GMing and such activities should be fairly easy.

The fine art of negotiation

When you win a conflict, you can suggest how the events unfold. Few key principles: Make it interesting to everyone. Don’t reach into the future, unless some character has oracular powers, and even then, giving traits is a better idea. Do give traits liberally. Try to give ambivalent traits that are not clearly good or clearly bad, as they are more fun to everyone. Give obscure and fuzzy traits.

Should some other player have an idea of what might happen, do listen to them. The idea might be good. Negotiation is preferable to dictation. But don’t reach into future.

In other words: Consequences should guide and influence future play, not dictate it. Traits are an excellent way of guiding and influencing, but not dictating.

Pacing

One of the actual reasons for having a GM is pacing. At the start of a session, it is necessary to open new story threads and expand on any potential for interest. At the mid-point, one should start focusing on the key story threads. At the end of play, only the key story thread should be there, others forgotten, for now.

The scope of traits

People should have a rough vision of what the traits in play mean. In game it is negotiated further. General guidelines: Only give full trait value in dice if the trait completely fits with the use. Almost always give half the trait in dice, as long as it can be justified. Stingy with full dice, generous with half dice.

Further development

The following are ideas or variant rules. They might be tested at some point and accepted as official at some point.

Death

A way for characters to die in play would be nice. Something definitely under player control, though. My tentative rule suggestion goes thusly: A character whose name is on the list is in conflict. Already has the player removed the character from the list to get bonus dice. Should the player want to do this again, the player must remove all occasions of that particular character from the list (permission of all players whose name is listed with that character is required). One die per name removed and the character may die as a consequence of this conflict.

Alternatively: 2 dice per name removed and the character must die as a consequence of this conflict.

Negotiation tweaking

One of the following…

  • Upon ending a conflict, the loser must suggest what happens and the winner can accept or deal harm.
  • As above, but freefrom brainstormy negotiation.
  • At any point in a conflict any player, or some specific (losing, winning, …) player can make suggestion. Anyone can answer by rolling more dice and describing more, or the suggestion may be accepted.

Things that don’t work as they should

Messy conflicts with several sides, especially if someone is throwing fireballs or other area-of-effect things. Generally, harm and messy conflicts.

Permalink 1 Comment

Hacking together a game

6 May, 2008 at 11:51 pm (game design, game element, persistent fantasy) (, , , , )

Vincent Baker a.k.a. Lumpley has published a game called In a Wicked Age. Being the cheap bastard I am, I won’t buy it (unless Arkkikivi/Arkenstone stocks it, at least), but will rather hack together something vaguely similar and play it.

What makes the process fun is that I have neither played nor read IAWA.

Components

The parts are, in no particular order, the List, the mechanics, the way resolution is used, and the random generator. Their implementation is explained after first explaining the components on more general level. And, as before, there’ll be one GM as a default assumption.

The list

This is stolen more-or-less directly from IAWA. Whenever a certain condition is met, the relevant character is added to the bottom of the list. Character can be crossed off the list by the player of that character to get a bonus. Whenever a game is played and the list is not empty, a number of characters from the top of the list are automatically in the game and their names are crossed off.

A (short) list might look like the following, with the character name first and player name in parenthesis after it. One entry has been crossed over. (Usually, there would be a huge swarm of entries crossed over in the beginning, but that is not very illustrative.)

  • Kisfal (Gastogh)
  • Ceosinnax (Tommi)
  • Kisfal (Gastogh)
  • Mori (Thalin)
  • Animagynth (Gastogh)

The random generator

The idea behind having a random generator is that at the start of every session/scenario/story/game (choose whichever is appropriate) a number of entries is generated and those are used to build the starting situation. I personally use Abulafia, but other generators can fit the bill. If one wishes to be independent of computers writing down or printing out a suitable list is advised. Number it, use dice or playing cards or whatever.

IAWA was what sold me to the concept of using random generators like this. (Actually, a random thread or two about IAWA, but the point remains unchanged.)

The mechanics

Characters are composed of a (finite) number of freeform traits. At least one should be an archetype or profession or something similar. Each trait has a numerical value, which directly determines how many dice it is worth in conflicts where it is directly and unambiguously applicable. Halve the number for somewhat applicable traits. (The idea of freeform traits is originally from Over the Edge; the numeric value corresponding to number of dice is from somewhere.)

When two characters are in conflict they get dice as above. Not all of the dice need to be claimed at once; it is possible and recommended to first roll whatever is most relevant and then add more dice from other traits if necessary. This bit stolen from Thalin‘s current victorian game, where it is not really doing anything due to there being too few traits per character. Any flaws give dice to the opposing side. If side 1 has no applicable traits, other sides have their pools doubled and side 1 gets a single die. Good luck.

Once dice are rolled and both sides as satisfied, or have run out of traits they intend to use, the dice are compared as per a method I have used before: First remove opposing and equal dice, then the side with highest remaining die is the winner, margin of success equals the number of dice that are higher than all the dice of the opposing side.

This didn’t really work in the previous incarnation, largely because there were too few dice on the table and I used too few dice for the opposition. The lack of a sufficient number of interesting traits also made it stale. Hopefully this attempt will work out better.

One should note that the resolution is very chaotic; it is possible for a single die to turn a minor defeat into a major victory. This is very much intended, so that one who is just about to win a conflict will be tempted to use all traits, even the ones that are of a somewhat questionable nature.

The resolution

After dice have been rolled (as above), the winning participant (player or GM) suggests what happens; the losing side either accepts that suggestion or takes harm equal to the margin of failure in the conflict. This, again, is from IAWA. The idea is that the winning participant needs to suggest something the losing participant finds interesting (or be content dealing harm, which won’t actually solve anything).

The resolution generalises to several participants: Whoever wins has a total margin of success that can be divided among the opposing sides. Every side with successes above the opposition can do this. All the dice can be targeted at single opponent or they can be divided in arbitrary way among the opposition that was beaten.

This we will play(test)

This is an explanation or example of play, which reveals details not included above. Assume everything written above still applies.

In the beginning

I mixed several appropriate generators on Abulafia to create the fantasy oracle compilation I’ll be using in this game. The oracle seems to generate too few actual characters; I’ll have to see if that is an actual problem. An example of output:

Ore which seems to whisper with incoherent voices.

The guardian spirit of a foolhardy, naive, reckless and impressionable young person.

A genius of flame, imprisoned within a brass mirror. (Might be a typo; maybe should be a genie.)

Forest of Eternal Peril

What is relevant is that there are explicit and implied characters generated. There’s the piece of ore or whatever resides inside it, if anything. There’s the guardian spirit and the foolhardy youngling. There’s the fire genie. And there’s whatever, if anything, that resides in the forest of eternal peril, whatever that is.

Part of the list may be ignored; namely, if a player is not present, all entries keyed to that player are simply ignored. If the list is empty (of relevant entries), every player selects something implicitly or explicitly generated by the oracle. If there is something relevant on the list, take half the number of participants, rounds down. This many different characters, counting from the top, are included in this session. The other players take characters implied by the oracle.

Assuming three players and one GM, the cast of player characters might be as follows, with traits and their values listed in parenthesis. Starting limitations: Up to three traits, up to six dice per trait. Scaling: 1 and 2 are minor, 3 and 4 significant, 5 quite powerful, 6 a bit too powerful to be used very often.

  • An efreet (genie 5, essence of flame 4, entrapped 3)
  • A kid (street kid 4, naive 3, “The stone guides me.” 2)
  • A guardian spirit (unseen 5, protect the kid 6, mute 4)

The starting situation could be: The kid, following the whispers of the stone she carries have taken the kid to the forest of eternal peril, where she discovered a beautiful brass mirror lying on the bottom of a pond. Her guardian spirit could only watch as she scrubbed it clean…

The next task is to determine something for the characters or the players to strive for. This can be formal (a trait) or informal, but the characters should bump into each other frequently.

The play

Characters done and the starting situation established it is time to play. Feel free to skip the next paragraph; it is mostly dry mechanics.

There’s the normal narration and roleplay and so forth until a clear conflict emerges; at least two entities, named or not, are in conflict more severe than mere discussion (arguments, intimidation, swindling, … are not mere discussion). For example, the efreet wants the kid to free it. Efreets are good at bargaining (that’s their purpose), so the efreet starts with 5 dice. The kid starts with 3 dice for the entrapment, the power of which makes it harder for the efreet to be released. Efreet: {6, 4, 3, 3, 2}, kid {3, 2, 1}. After putting the matching dice aside, one is left with {6, 4, 3} for the efreet and {1} for the kid, with {3, 2} aside from the efreet and the kid. The guardian spirit protects the kid from the vile efreet’s influence: 6 dice for protecting the kid. Dice show {4, 3, 3, 2, 2, 1}. Since the kid and the guardian are allied and not the same entity, one 3 and one 2 the spirit rolled is set aside, as the kid and the efreet also lost one of both. Then another 3 and 4 are cancelled from both the efreet and the spirit. This means that the relevant sets are efreet {6}, spirit {2, 1}, kid {1}, aside several (which I won’t write down; this is a lot easier when there are actual dice on an actual table and they are moved and grouped). To make matters worse, the kid is naieve, which the efreet’s player can draw upon, giving extra three dice, which show {5, 4, 2}. Other traits are not claimed, so the final result is efreet {6, 5, 4}, guardian spirit {1}, kid {1}. The efreet has 3 successes over both opponents, the others have none. The dice are biased tonight.

So, the efreeti has total 3 successes over the kid and the kid’s guardian spirit. Efreet’s player offers that the efreet is released from the mirror, owes the kid three wishes, but the kid does not know that with mere 1 die to back it up. To the other involved player the efreet’s player suggest that the efreet can see and interact with the guardian and does not seem an immediate threat, backed with the remaining 2 dice. Both suggestions are cordially accepted. Efreet’s player changes the trait “entrapped 3” to be “Those who imprisoned me shall burn! 3”, which seems appropriate, so the GM and other participants accept. There would naturally be some roleplay involved in describing these events.

What if one of the players had not accepted the suggestions? Their characters would have taken 1 or 2 harm (kid and spirit, respectively). The meaning of harm: One can only use traits with value exceeding the total harm suffered. That is: Harm 4 and only traits with 5 or more dice can be used. This does not affect opponents using weaknesses, but does affect the harmed character exploiting the weaknessses of others. Harm is recovered only when the session/story ends, and is then recovered completely. Harm equaling or exceeding the character’s highest trait (or all traits, same thing) implies that the character is unable to do anything meaningful; maybe dead, maybe imprisoned, maybe searching for more peaceful lands elsewhere. Such characters, if they are on the list, can be encountered later.

The list, right. Current idea is that any character losing a conflict gets on the list. This condition may be too lenient, but only play(testing) will tell. More restrictive conditions in the same spirit: Only when when actually suffering consequences for losing a conflict (marginally more restrictive), only when losing a conflict and accepting the interesting consequences suggested by the other participant (as opposed to taking harm; if people take harm too often, I’ll implement this), only when taking harm (feels too limited and encourages taking harm, which I assume will not be that interesting). Crossing the name of the character you are currently playing has the following effect: If the name is on top of the list (of the characters whose players are present), get 3 extra dice. This is typically a very significant lucky incident or divine favour. If crossing the name on the bottom of the list, get 1 die. For any other location on the list, get 2 dice. This can be done exactly once per conflict per character. These dice are not restricted by harm. Alternatively, the player can choose to cross over all places where the character is on the list. This gives single die per name, and hence should not be used unless there are at least three names of that particular character present. Note that this has a significant chance of permanently removing the character from play. Take care, use wisely.

In the example, the guardian spirit and the kid get on the list. I think their order will be first the spirit and then the kid; this because the spirit risked 2 harm. If this is not sufficient to determine the order, remaining draws are handled by the GM by pure fiat (which may include asking the players if they have preferences).

A character can get on the list if and only if the character is named.

Character change

When participant feels a character has changed in some significant way, he ought to tell that to the other players and any relevant change in traits should happen immediately. Training is suitable. Saying the character has practiced something during his or her downtime is likewise suitable.

One of the more interesting possibilities is trait change due to losing conflicts: The winning participant may suggest changing, adding or removing a trait. For example, an assassination attempt could lead to traits like “crippled”, “wounded”, “They all are out to get me!” or “nervous”.

The end

Game master gathers character sheets and the list. They are persistent from session to session. Any detail generated about the game world should likewise be recorded somewhere, because emergent fantasy setting are fun and useful.

A note on design

This is very much bricolage-style design; that is, building from old parts, mixing them together and hoping they interact in good ways. The purpose is to create a game that I can play with, well, anyone, even if the groups of people change, there is irregular attendance, or otherwise separate groups are brought together in, say, Ropecon. Episodic gaming, pretty much.

Permalink 1 Comment