Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
January 17, 2018
arrowPress Releases

If you enjoy reading this site, you might also want to check out these UBM Tech sites:


Enabling Deeper Relationships Through Multifaction

by Bart Stewart on 02/05/10 01:46:00 am   Featured Blogs

2 comments Share on Twitter    RSS

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Among the numerous good points in her Gamasutra blog post Touching Players Emotionally, Shelly Warmuth wrote the following:

In the final analysis, the answer would seem to lie in creating a game that offers small ethical dilemmas.  The player should be making decisions with little thought to the outcome.  The decisions should seem so insignificant that they are invisible to the player, making them unaware of how they are affecting the game.  And yet, the decisions should be significant enough to affect the player in a small way even when making them.

On reading this, I was reminded of some similar thoughts I had a few years back. Then and now, I agree -- the gameplay mechanic of small individual decisions that have limited meaning when taken, but that accumulate over time into a significant history, is an opportunity for interesting dynamic storytelling that deserves to be better explored.

I believe the MMORPG concept of "faction" offers a good starting point for generating in-game social environments that carry consequential weight. Implemented more enthusiastically than in today's games, we could find our gameworlds becoming vastly more dynamic and interesting through a feature I'm calling "multifaction"... but first let's consider what's meant by "faction" as the term is currently used.


"Faction" in massively multiplayer online roleplaying games (MMORPGs) carries several meanings, all of which are related to determining how non-player characters (NPCs) who are members of in-game organizations tend to react to the player character. For each NPC organization (such as a Warrior's Lodge or Starfleet or the Rebellion), a number is stored in the game's database which represents how that organization regards the player's character. A high number usually indicates that NPCs belonging to the organization will react favorably toward the PC, while a low number tends to mean that NPCs belonging to that group will react negatively toward the PC.

This leads to a number of loose usages of the word "faction." For example, in the recently-launched Star Trek Online, there are two factions to which player characters can belong: Starfleet and the Klingon Defense Force. So a player might say, "I'm with the Starfleet faction." Another use of the word reflects the player's understanding that faction is a gameplay mechanic: "I need to run some missions to build up KDF faction" would mean that the player wants Klingon-aligned NPCs to react more favorably toward his character, so he intends to take in-game actions that he knows will increase the magic number of the KDF's "faction" that determines how Klingons will see his character. So "faction" can refer both to an in-game organization and to the numeric value associated with a group that determines how NPCs belonging to that group regard the player's character.

(Note: faction is different from Karma as used in Fallout 3 or Renegade/Paragon scores in Mass Effect in that those good/evil measures are stored solely "inside" the player's character, rather than in NPC groups. Another way of saying this is that karma and Paragon/Renegade scores dictate the reactions of individual NPCs toward the player's character, as opposed to faction which measures the regard toward the player character of entire organizations of NPCs.)


The important thing about "faction-as-a-number" is that it constitutes a form of memory. Numeric faction is how an NPC group "remembers" the actions you've taken that affect the supposed goals of that group. And the great value of this simple mechanic is that it allows players to engage in relatively small actions that, individually, have only a marginal effect on any group's numeric faction... but enough such actions over time can add up to produce strong reactions. Rescue one person and you earn a small reward -- rescue hundreds of people and you can be elected Governor. What you do over time helps to create your particular story.

In most MMORPGs that's as far as it goes. Earning a high faction with the two or three player-joinable NPC organizations that the game's developers implemented can net you some nice goodies. In a few other MMORPGs, however, it's also possible for each player character to earn or lose faction with multiple NPC-only organizations. This creates a very simple form of dynamic story-telling: player characters who've earned differing amounts of numeric faction with an NPC group are treated differently by that group. Through his freely-chosen individual actions one player's character might gain high faction with some NPC group, causing that group to reward his character with some benefits. At exactly the same time, another player might choose to repeatedly perform actions that run counter to that NPC group's goals, causing NPCs belonging to that group to react with outright hostility toward the player's character.

Faction, then, is a means through which small individual actions can eventually accrue great meaning in the gameworld, with significant consequences to a player's character in both gameplay and story. There's no single great act of heroism or cowardice that instantly determines the nature of the gameworld. Any such act is gameable; the player can go online and read what to do in a particular scenario to force a particular outcome. With faction, what matters instead is the consistency of your behavior over time.

In short, faction is a way to design a single game whose social element evolves differently for different players based on their actions. Player A enjoys a world in which he is welcomed almost wherever he goes; Player B must skulk from place to place to avoid the hatred of all; Player C is universally loved in some parts of the world but would be killed on sight in others... and it's all the same game that got shipped.


This is a pleasant vision. But I'd like to propose something even more interesting. Specifically, consider the effects on a massively multiplayer gameworld of implementing the following two rules:

1. What if individual actions by player characters were defined to produce changes to multiple factions?

2. Instead of NPC groups only storing numeric faction for each player, what if NPC groups could have faction for each other... and player character actions could affect those factional values?

Let's consider an example of applying these two rules as gameplay in a MMORPG based on Star Trek. Suppose that your character, as a Starfleet officer, saves a stranded Romulan NPC from death. Your faction with several NPC groups is already high enough so that you're considered a member of those groups. The Romulan, in turn, belongs to several other NPC groups. What kinds of factional changes could your saving a Romulan officer's life produce among all these individuals and groups?

Here are some (but by no means all) possibilities that could be predefined to be applied when such an event occurs:

  • You gain 50 "personal faction" with the individual Romulan NPC you saved.
  • You gain 3 faction with Romulan Fleet Command.
  • You gain 5 faction with Starfleet Command.
  • Starfleet gains 2 faction with the Romulan you saved.
  • Starfleet gains 1 faction with Romulan Fleet Command.
  • You gain 10 faction with the Federation Council.
  • Starfleet Command gains 2 faction with the Federation Council.
  • You lose 3 faction with Starfleet Security.
  • You gain 3 faction with Section 31.
  • You gain or lose some amount of faction with your NPC commanding officer based on the strength and direction of his/her feeling toward Romulans.
  • You gain 20 faction with the civilian Federation-Romulan Benevolent Society.
  • You lose 12 faction with the civilian Terra Insolitus movement.
  • Your personal faction with the Klingon Empire changes by 1/10 of their current faction with the Romulan Star Empire.
  • The Romulan you saved loses 10 faction with the Tal Shiar.

Now take your individual action and add it up with all your other actions over time... and then add all the actions of all the other players on your server... and then add up all these actions over several weeks and months and years of play.

That game world gets a bit more interesting, doesn't it? With thousands of players taking many small actions in an uncoordinated way over time, the state of the gameworld becomes unpredictable in a remarkable way. To continue the example of a Star Trek MMORPG suggested above, consider: what happens to the Federation's relations with the Klingon Empire if so many players rescue Romulan NPCs that the UFP and Romulan Star Empire eventually have high faction with each other? What if a lot of Starfleet characters started taking actions that led to Starfleet and the Federation Council having negative faction with each other?

And that's what I mean by "multifaction," and why I think it's being grossly underused in massively multiplayer RPGs.


This brings me to an obvious objection: in gameworlds based on licensed properties, certain groups simply would never act at cross-purposes to each other -- for them to do so would break immersion in the fictional world of the license. So a multifaction gameworld of the kind I've described here would probably, as a practical matter, need to be able to set upper and lower bounds on factional levels of certain groups toward each other. It might be interesting if Starfleet and the Federation Council didn't like each other for a while, but letting them go to war against each other would be too radically different from the desired kind of world. Upper and lower fenceposts would prevent such outcomes without eliminating the possibility of significant positive or negative regard of groups for each other, with the attendant friendly or hostile behaviors that would result.

Another objection is that having to define many possible factional changes for every action a player character could take would significantly increase the amount of development work necessary (for a type of game that generally already has more nice-to-haves on paper than can be implemented in any reasonable amount of time). There's some justice to this objection. It would be more work than simple faction requires.

However, given than the effects of all player actions need to be defined as code anyway, it is not unreasonable to think that a MMORPG development engine could be designed such that, along with visual effects and chain rule firings that can be associated with any player character action, factional effects could be one more type of association to fill in. There'd still be some work required to define the multifactional effects of each action, but a well-constructed game development toolkit could take some of the drudgery out of defining multiple factional effects for each player action.

Yet another objection is one leveled at some MMORPGs today: what if a player performs some actions that cause an NPC group to "hate" his character? This effectively prevents the player from ever being able to experience the beneficial gameplay effects of that particular character having positive faction with that group -- they hate that character forever.

Happily, multifaction doesn't make this problem worse; it actually provides a solution. In a typical MMORPG a player character must interact directly with an NPC of a certain group to gain positive faction with that group, but that's impossible when the character has lost so much faction with that group that NPCs belonging to it will no longer talk to the player's character. In a multifaction gameworld, it's possible to perform actions for other NPCs that indirectly raise faction with some other group. To put it another way, suppose that through some series of actions your player is hated by group A, while group B is neutral toward the character. In addition, suppose that Group A has significant positive faction toward Group B. As your player takes actions that Group B finds favorable, raising their faction toward you, Group A will begin to increase their faction toward you because of their positive regard for Group B.

In a sense, multifaction allows players to be measured not solely by their own actions, but also by the reflection of their actions toward various groups. In a multifaction gameworld, you're known by the company you keep. What you do matters.


And to sweeten the pot in one more way, there's another virtue to implementing a multifactional gameworld that's worth considering: in a gameworld where hundreds of groups have varying levels of regard for each other, diplomacy becomes a viable form of gameplay.

If groups have numeric faction toward player characters and toward each other, and player character actions can affect those factional values in ways that are individually small but add up over time, then that generates opportunities for players to perform actions that are intended to alter the factional regard of NPC groups for or against each other. In other words, when groups have high or low regard for each other and players can slowly affect those levels of regard, that creates the necessary foundation for diplomatic gameplay mechanics.

Diplomacy is particularly interesting because it's non-lethal competitive gameplay that players can engage in on behalf of their respective player factions and against each other. Because it hasn't been strongly implemented in any MMORPG (of which I'm aware, anyway), it's a feature that a new MMORPG could claim as a uniquely distinctive selling point.

Still, its novelty also suggests it would benefit from being presented to players in a way with which they're familiar. Diplomacy in a multifactional gameworld is probably best understood in light of two other existing mechanics of MMORPGs: aggro management and PvP (player versus player) play.

Diplomatic gameplay would be defined as needing three characters: your character, another player's character, and an NPC representing a group whom both of you are trying to sway to the side of your character's home faction. Each player character would take turns engaging in various diplomatic maneuvers -- threats, promises, accusations, rewards, innuendo, flattery, insult, etc. -- which act like /taunt or /invisible to affect the "aggro" of the NPC. These actions would be represented in some format appropriate to the milieu of the gameworld.

Your player opponent, meanwhile, would be doing the same kinds of things, and the player who more skillfully manipulates the NPC would win that encounter. Diplomatic engagements would thus feel very much like indirect PvP combat. (And isn't that a fairly good description of diplomacy?) I'm not a fan of "aggro management" as tactical combat gameplay (as noted in "An Alternative to Aggro"), but I think implementing it as the cut-and-thrust of diplomacy feels completely natural -- diplomat players really would be trying to modify the internal state of an NPC.


To sum up, then: in a multifaction MMORPG, the NPCs and NPC organizations in the gameworld affect and are affected by each player's actions in combination with those of other players and accumulated over time. This not only layers a social terrain onto the gameworld that is constantly shifting to keep the gameworld surprising and fresh, it enables non-lethal PvP gameplay that supports one's home faction.

You might even say that the functional friction of fictional faction generates factional fiction.

Or maybe you wouldn't.

At any rate, comments regarding this multifaction concept are welcome. I hope there'll be a few that see some merit in the idea, but thoughtful objections will be appreciated as well. Bottom line, I'll be content if this post stimulates some thinking about how the faction mechanic can be put to better use in games.

Related Jobs

Respawn Entertainment
Respawn Entertainment — Chatsworth, California, United States

Senior Character Artist
Qualcomm — San Diego, California, United States

Game Designer
Sanzaru Games Inc.
Sanzaru Games Inc. — Foster City, California, United States

Environment Artist
Sanzaru Games Inc.
Sanzaru Games Inc. — Foster City , California, United States

UI Artist

Loading Comments

loader image