Classification of archetypes of RPG characters

Started by
5 comments, last by Trapper Zoid 18 years, 7 months ago
Forgive me in advance if any of this is unclear. Part of the problem here is one of terminology, so I find it hard to express exactly what I mean on this topic. The Short Description: I'm looking to come up with a classification hierarchy system of the archetypes of RPG characters, and how they relate to each other. I'm starting this thread to brainstorm on the best way to build such a classification system so it's as complete as possible. What are your thoughts on this? I'm open to any brainstorming ideas from anyone! Some random brainstorming on this from me... I've been working on systems for automated interactive storytelling for a while, and one of the most important elements of such as system is the characters. For my present game design, I need to make a classification system of the archetypes of RPG characters. However, from what I've read and seen there's always a blurring between character roles. What I'm aiming for is a decent hierarchical classification system that covers most of the character types that are present in RPG stories (at least the most important character types); similar to the inheretance systems present in object orientated programming, a directed graph-like structure where one higher order character type can take on the properties of more than one lower order character type. I'll try to explain myself in more detail later. I've read various classifications of characters by other people for different mediums. Propp has defined for Russian folk tales character archetypes such as the hero, princess, villain, donor and helper. Campbell has archetypes such as the hero, mentor, threshold guardian, herald, shapeshifter, shadow and trickster. In these systems these archetypes are more like roles (so a character can shift between different types), but they are also a bit vague. I'm finding it difficult to conceptualise a good way to apply this to the domain of RPGs. Let's take the most central character in RPGs, the hero. Most often, the story completely revolves around the hero (as that's the character the player is controlling). However, in most RPGs there's more than one hero. Sometimes there's a chief protagonist (a "lead" hero), however sometimes there's a core group of leads. Often there are playable characters in your team that are still "heroes", but do not have as much relevant to the storyline as the lead protagonists; in essence they are "lesser heroes". Does that make them a slightly different class of character? Then there's the antagonists, the villains. Again, there's usually a chief villain, and often a "fake chief villain" who acts a bit like the real chief villain until the true villain is revealled. Then there's the villains "lieutenants", who still act a bit like villains in that the foil the heroes' plans, but seem to be a separate subclass. However these lieutenant types often also cross over into the ranks of the heroes, or were once heroes before they fell into their role of villain. Of course, there's also the common garden variety NPC; the common villager, who usually is relegated in RPGs to saying simple phrases, or selling things. But some of these villagers have a more active role in the story, in positions such as the village elder, chief or monarch, where they start story elements for the heroes or do actions on their own (and thus inheriting properties of the hero). The point of this is how many unique character archetypes are there? Could it just be simplified down to something like common NPCs, heroes and villains (three archetypes)? Or even down to just common NPCs and heroes, with villains being "heroes" from another point of view (this is a representation I'm increasingly leading towards; with NPCs being passive agents in the story and heroes/villains being the active agents). Anyway, this has rambled on for long enough; I'm interested in your thoughts on how the different types of RPG characters can be classified.
Advertisement
Re:
Classifying Characters for a Game Story Engine

It is easier if you know what qualities your engine plans to manipulate. In terms of the Terrain, there are three classes of attributes pertaining to a character:
    Resource - The amount of resource the character has, able to generate, able to obtain, and able to transfer. In the context of a story, the main form of Resource is information. Most NPCs are this type. For example, a generic shopkeeper is a character with infinite resource content (of the items it sells), and indiscriminate rule of resource transfer (as long as your char has enough money it sells).

    Mobility - The ability to travel on the Terrain. The Terrain is the space where the abstract path between the initial state and the goal state lies. It is not the game map. Mobility is the ability (and related attributes) to cause state changes pertaining to the Terrain. In general, a Hero has high Mobility.

    Influence - The ability to shape the Terrain. For example, a character that can define the goal state, introduce obstacles on the Terrain, or add/remove the paths availble on the Terrain. A villain is an example of character with high Influence.


* These are classes of attributes. A character can be scored on these attributes. The engine reads the variables under these classes for simulation.

Classification of common archetypes based on these attributes:
(From the perspective of the Hero in the context of a simple RPG)

    Hero - High Mobility character, also high in Resource transfer abilities. The Hero is usually the only one in the game that initiates state changes.

    Princess - In the 'save the princess' type stories, the Princess on the Terrain means nothing. It has almost no values in any of the attributes. It is like the flag in golf marking the hole. The flag that you remove when you get close to the hole.

    Villain - The major I-character. In most dungeon RPG, the Villain obviously moves, but movement in the gameworld does not imply mobility. In terms of the State Terrain, the Villain is immobile. The Villain is the medium (or agent) through which the Engine shapes the Terrain. From the perspective of the Engine, the Villain is its ammo.

    Donor - Information or Resource giver, also with ability to Influence. The old guy that hints you a secret shortcut is an I-type character. The sword master that teaches the Hero new skills is can be an I-type or R-type character, depending on whether you define the specific skill as a path or a resource.

    Helper - Resource and Mobility. Shares Mobility with the Hero.
    Mentor - A kind of Donor, R-type. Provides
    Threshold Guardian - A kind of Donor, I-type. Opens state transition paths.
    Herald - I-type. Define/changes goal states.
    Shapeshifter - I-type, removes/changes goal states.
    Shadow - I-type, removes state transition paths.
    Trickster - R-type, gives you bad Resources or false Information.


The attributes are defined in the axial level, that allows you to classify the archetypes in the 3D archetype space. The purpose of the classification is for the Engine to recognize what it can do that is 'appropriate' to what the character would do. Primitive example:

Define characters:Hero: M  	  // The Hero can MoveVillain: I- I+    // The Villain can Influences the Terrain negatively, 		  // and define the goal state Shopkeeper: R+	  // The Shopkeeper can give resourcesSage: I+	  // The Sage can gives hints and reveal paths

In this case we don't include R- characters because the Terrain takes up Resources when the Hero makes a move. When the Villain creates an obstacle, it makes the Hero more likely to choose a path that consumes more resource (unless the Sage tells the Hero another way out). The Villain does not drain resource directly.

Processes:

1. Suppose the game begins with the Hero knowing absolutely nothing, the first task of the Engine is to pick an Agent that can hint or tell what the Hero is supposed to do. In this case, the only choose is the Villain. So the Engine picks the Villain, and through it defines a goal state. If there are more candidates the Engine has to resolve what goal state to be set.

2. After the goal state is set, the Engine activates the corresponding nodes that will inform the Hero about the goal. In this model, the Engine does not just present the goal. The Hero has to has to bump into a node that presents it. If the Villain chose one out of three of its evil plans A, B,& C, you can imagine the Hero starting the game surrounded by Information landmines marked A, B or C. When the Villain selects a plan, the corresponding mines are activated. The landmines are distribute in a way such that the Hero will run into at least one. Since the Terrain is a visualization, the mechanism translates to schedule an informative event subjected a timing requirement.

3. After the Hero hit a mine (or a roadsign), the Hero starts going toward the goal state. In this simple case, the state roadsign is correct. From this point on, the task of the Engine is to repeatedly select an Agent that shapes the Terrain. In this case, the only one that can shape is the Villain. The Engine picks it, and from its inventory of evil ideas, picks a state obstacle.

4. The Hero faces his first obstacle, Resources are used. In order to continue, the Hero needs to refuel. In general, the Hero can go back to town to refuel, but that is kind of boring. Depending on how well the Hero fought, the Engine will pick a help to refuel the Hero. In this case the Engine picks the shopkeeper. A wandering shopkeeper. Hero gets refueled and makes a new friend, and the story continues.

5. From the Villain's inventory the Engine picks another Terrain deformation. This is the second encounter, so the event is predictably more intense. Resources is used. Engine picks someone or no one to refuel the Hero. This cycle continues, until...

6. As the story continues, due to the Hero's choices, the Hero will have different Resources and be at an arbitrary location on the Terrain. It is not guaranteed that there is a path from any state to the goal state. Or maybe there is a path by the Hero has no Resource to take the path. These are deadends, gameovers. It should be obvious that a common intelligent part of playing a game is to avoid such deadends. For a bug free Engine, the Hero can arrive at a deadend for two reasons: being stupid; or being desperate. There are situations when a player will make a Hero do something risky. Taking risk should be encouraged because it directly correspond to emotional response. If the Engine detects that the player is in a courageous risk taking mode, the Engine rewards the player with additional paths. (Engine picks the Sage and it tells the Hero about a way to defeat the Villain that is also risky) In this philosophy, 'the secret paths' are not free passes, they are additional challenges matching the increasing Intensity. The idea is to detect how much risk the player can tolerate genuinely. When there is limited I+ paths provided by the I+ Agents, the Engine needs to monitor the health of the I+ possibilities to preserve the Arc.

These are my thoughts on how archetypes can be analyzed based on how the components are used for the Engine. You might have your own set of requirements for what your Engine needs to do. From that, you will get a different set of features to classify characters.


[Edited by - Estok on August 9, 2005 3:48:27 PM]
Thanks, Estok. I'd thought my post was too confusing to get any replies!

Quote:Original post by Estok
These are my thoughts on how archetypes can be analyzed based on how the components are used for the Engine. You might have your own set of requirements for what your Engine needs to do. From that, you will get a different set of features to classify characters.


That's pretty much what I've thought; the other systems that other people use won't be applicable to what I want to do. I've been playing with ideas like yours for automated storytelling systems, based on Propp or Campbell's model of characters. However, for this project I'm wanting the system to emulated the stories found in present console RPGs, so I'm wanting to abstract away the character types to get their core archetypes. While the Propp and Campbell models are similar, I don't think they're quite the same (for example, manifestations of the "Princess" archetype are more of a passive reward object in Russian folktales; in RPGs the love interests usually have an active role in the story).

I have some idea about the archetypes I'd like; normal characters, heroes and villains are the three that I'm playing with at the moment. I think I'll post some more focused questions on those archetypes when I get stuck, like the one I've got running on villains at the moment.

Well by archetype do you specifically mean hero/villan/npc or are you trying to get combat style archetypes (mage/tank/ranger).

generally speaking you have ,in most stories, main characters then a cast of supporting characters. Then those characters fit into roles: main hero, main villian, damsel in distress. then the roles lead to archetypes of characters that fit into the given role but add specific flavor to that role as to make them memerable. you can have the "amnesia hero" ala cloud strife, or you could have anti heros that just dont care( cant think of one off hand).

But basically thats how it goes: Generic character profile(main/supporting/minor etc) -> character role -> character archetype -> unique character traits.

hope that helps some.
Thanks vaneger, I'd somewhat forgotten about this thread.

Your breakdown is pretty much what I came up with. Characters will be classified by their story importance (as you wrote): lead, supporting, minor, and "extras" (i.e. those NPCs who say the same thing over and over again [smile]). Then they have role (hero or villain), and some personality characteristics. Yes, it seems as if we've come to the same conclusion here, which is good; that suggests that the breakdown I've come up with is a good one.
I think Dramatica has an interesting system - it classifies characters by their roles in the thematic argument, which fits very nicely with the vector theory of plot since each character is supposed to embody a thematic vector. In dramatica the roles are:
Protagonist - proactively pursues the 'right' goal
Antagonist - the antagonist proactively pursues an opposing 'wrong' goal
Guardian - exemplifies the 'right' method
Contagonist - pushes the protagonist to use the 'wrong' method

Any one of these can be your viewpoint character or Main Character, and the one with which they have the most conflict then becomes the Impact Character (which determines whether the main conflict is an issue of goal or method).

I want to help design a "sandpark" MMO. Optional interactive story with quests and deeply characterized NPCs, plus sandbox elements like player-craftable housing and lots of other crafting. If you are starting a design of this type, please PM me. I also love pet-breeding games.

That's some good points, sunandshadow. At the moment in my design I've been focusing on the importance of the character to the plot, as that is somewhat important to the design balance at the moment. I'm also going to have their function in terms of the story, but at the moment I've only considered having heroes and villains. This is a bit simple having a black-and-white split of characters, but since the generated stories in my game are going to be simple anyway it's a good starting point.

It's a little bit tricky to use the standard dramatic techniques as there are multiple heroes and villains; it's a bit like a comic book universe in that respect. I'm designing the game to play like a city management game or the Sims, so the player isn't actually in direct control of any of the characters. So terms like "protagonist", which imply the one most important hero, can't really be used.

But thinking about this, I should be pairing off heroes to villains, so that I can use the standard story techniques in my game. That would work quite well, giving each character the potential to have a nemesis.

This topic is closed to new replies.

Advertisement