12 October 2009
MMO Design: Rich Faction
Reposted on new blog at http://dynamitochondria.blogspot.com/2009/10/mmo-design-rich-faction.html
Faction is a widely used concept in persistent-world games, especially in the MMO space. Blizzard's World of Warcraft is the best known example, using Faction in two related contexts. When you are first creating your character, you choose Horde or Alliance, and that's your Faction. It simply specifies your "side" in the ongoing conflict featured in the setting. However, Alliance and Horde are also examples of a somewhat more complex mechanic. Each of them has a numerical value representing how well you are "liked" by that side.
In addition to these big two, there are any number of smaller factions, ranging from small towns to large organizations neutral in the Horde/Alliance conflict. You increase your score with a given faction by doing quests assigned by that faction's quest-givers. The Faction award for a given quest is orthogonal to the XP awarded, so it's possible to earn Faction score for a mission so far below your character's level that it awards no XP. I've frequently heard this described as "Grinding Faction".
So what good is it? As you increase a Faction score, the NPCs associated with that Faction will change their behavior toward you, maybe giving you a quest that was previous locked, maybe giving to access to some unique resource such as special mounts.
That's interesting enough, but it could be so much more. For not much more complexity, and a trivial increase in the data storage requirements per character, Faction could be strongly enriched.
In the real world, one person's opinion of another person can't be summarized in a single integer score. Think about it. You have a friend you like tremendously but wouldn't trust to borrow the car or feed the cat during a vacation. You don't necessarily like them any less, but you have a low estimation of their competence.
Think Gilligan here. Gilligan has a high "like" score with the Castaways Faction but a low "estimation" score. They all know he'd go through hell to help them, and they'd do the same for him, but they're not fooling themselves about his competence, to the point that he gets underestimated now and then.
How can we apply this idea to a persistent-world game? Ordinarily, when completing a quest for a Faction, your single-value score goes up by a fixed amount. What if the Faction logic kept track of your defeats while attempting the quest? If you die three times in the quest instance but eventually get it done, you get the expected award to your Like value, but your Estimation value drops somewhat. In the future, the Faction's quest-givers assign you slightly less challenging quests that grant a little less XP. As your character becomes more powerful and your skill as a player increases, you find yourself getting defeated less, and your Estimation score goes up. You get harder missions with better XP. By adding one integer value per Faction per character, you create an automated adjustment to difficulty progression.
It makes a certain sense to track this data per Faction, since the types of enemies you face for different factions are often very different, and in-story different groups may not share information. But what if they do? How about a tiered Faction system that tracked the relationships between different Factions. The Blue Wizards and the Red Knights are both allied with the Purple Kingdom. If you do very well on quests for the Blue Wizards, your Estimation score goes up with the Purple Kingdom and the Red Knights as well. Maybe your Like scores go up as well, or maybe down. It makes sense to adjust just about any Faction value based on Faction changes with an associated Faction. The Blue Wizards are closely allied with the Purple Kingdom, but are in bitter rivalry with the Red Knights, who are also closely allied with the Purple Kingdom. You successfully complete a grand quest for the Blue Wizards and are awarded Like bonus with the Blue Wizards, a lesser Like bonus with the Purple Kingdom, and a Like penalty with the Red Knights. You can try to maximize your Like score with one Faction over the other or try to play both sides for some optimum aggregate Like score.
By placing Factions into a hierarchy with defined relationships, you've added interesting choices for the player. Faction is currently oversimplified and underutilized. Making only these two changes adds two very interesting dynamics to a persistent-world game. Enriching this simple attribute further could add unguessed possibilities.