Pro-Social Design

The question arose last week: how do you design around/against people being idiots and jerks? “You can’t fix stupid.” There is no 100% solution, because some people really are that dumb and others will go to great lengths as griefers, but there are better and worse designs in terms of the behavior they reward. If the system rewards pro-social behavior, it promotes harmony. If the player must make sacrifices to help others, you will see destructively selfish if not predatory behavior. Economics in two words: “incentives matter.”

For example, consider Marks of Triumph in The Lord of the Rings Online™: Shadows of Angmar™. The epic quest chain is a big feature for LotRO, but it was punctuated with instances that demanded full groups. If most of the population had completed them all, how did newer players and alts get through the epics? You asked someone to repeat one. Repeating one was a way to help friends, but you got jack for it. Your friends had to give something up, and you would not meet new people unless someone was a very charitable stranger (or, lucky day, you find a few people who need it, a couple of whom have charitable friends). Game update: repeating one of those instances began to award (once per five days) a Mark of Triumph; accumulate several Marks to barter for various rewards. The rewards were rather nice for when they were released. Pro-social behavior increased.

Because of how Marks were awarded, you did not need someone new to repeat the quest. This has the further benefit of letting you repeat older content without completely sacrificing character advancement, and developers want players to pay for recycled content. The downside is that it is more efficient to get a level-capped group and cycle through all the Mark instances rather than actually helping near-cap players on their first run-through. On balance, however, Marks increased pro-social behavior more than they inhibited it.

Looting is a major source of drama and anti-social behavior. Did you know that you cannot ninja-loot in City of Heroes? Not that there is much that you could ninja-loot, but there are not even corpses to loot. Rewards just instantly appear in each character’s inventory. So one person randomly gets the reward from the raid boss? No, everyone gets one. There are drawbacks to every system, particularly given the adjustments necessary to adapt it to another game, but there are ways around all the problems you assume.

Consider the related problem that there is nothing for you in that instance. Alice wants to run the caverns for her tanking trinket (best in slot!), Bob needs the mountain for his shoulders, and Carl is grinding crabs by the lake in case one drops the resource he needs. They could help each other, but that involves giving up their individual progress: pro-social behavior requires an individual cost. Or we could set up a different system for how you get “best in slot” that will not give people competing demands, such as a token/badge system that lets you adventure anywhere and then exchange for your rewards. Of course, that will encourage grinding in the most time-efficient area (drawback).

LotRO has an interesting mix on those loot systems. The top-tier armor involves tokens that you can earn in sets of dungeons, so you are not bound to any one. Freedom! When Mines of Moria™ was younger, you saw people who would never go to instance X again because they had their radiance pieces from there. So the armor rewards are mostly solved, and the weapon rewards are on a similar system, but all the other slots are on the same system in which the best tank ring drops here and the best DPS bracelet drops there. Is that a compromise between pro-socialization and encouraging people not to focus on the most time-efficient dungeon?

Consider the DPS meter. On one hand, it can be useful in assessing your performance and seeing where the group could improve. It frequently becomes its own mini-game at the expense of good tactics, with DPS encouraged to deal as much damage as possible to as many targets as possible as quickly as possible. Sometimes that works well, sometimes that causes wipes, and the tanks and healers frequently want to choke people. Also consider the gap between what is easy to measure and what helps the group.

Consider achievements for doing various absurd things. Someone on your Team Fortress 2 server is probably trying for one at any given moment. That Medic is not healing anyone because he is hunting for Scouts. That Heavy is just standing in a corridor spamming G in the hopes that someone will walk into a taunt kill. You see similar things where there are rewards for suffering; I left characters intentionally getting chain-stunned overnight for City of Heroes badges, and people also farm damage there. You get perverse results if the tank equivalent of the DPS meter measures damage taken and effectively punishes having high defenses (damage prevention is not measured).

Consider achievements for working together. The best of them will be for normal cooperation as opposed to situations you would perversely work to create. The TF2 Engineer achievements have lots of good examples. Identify how you want people to work together and place a bounty on it.

We could also consider pro-social design from the group perspective, rather than the individual. Consider how FFXI approaches (approached?) level ranges in groups. Everyone needed to be in a very narrow band, or else there were severe penalties. “Ding!” “Grats! /kick” Consider how CoX approaches level ranges in groups. Everyone is automatically the same level, which is exactly the right level for the instance. Group with anyone anytime: perfect. (We have discussed the drawbacks of the CoX instancing/scaling at length, and we will surely do so again.)

Really, pick any mechanic that sets the interests of the individual against the interests of the group, rather than aligning the two. People respond to incentives. If people are rewarded for leeching, stealing, or griefing, it will happen. Yes, some people will always be schmucks, but it will happen a lot less if you do not pay them for it.

: Zubon

5 thoughts on “Pro-Social Design”

  1. I think one of the root issues is the carryover effect of getting continually stronger from RPGs. It works in the solo game because you never need/want to go back to level 1 to bash goblins, but in an MMO (especially a more worldly one) you SHOULD want to visit all of the areas at all times, not just progress through from A to B to C to (expansion) D.

    Not to say remove progression, but don’t directly link progression to content going ‘obsolete’. Marks is one solution, but I’ve always liked games that allow even top-end characters to get some value out of ‘starter’ stuff, like in EVE where you can break down items into mats. The total gain is slower, but you are not ‘wasting time’ helping someone out, or doing something that is not the bleeding edge of optimized grind.

  2. Actually, FFXI did finally somewhat solve their level range issue. They added a feature called Level Sync, which means that you can limit the party’s level to match a specific member. Usually you sync to the lowest-level person so everyone is the same level. You still gain experience while synced, and gear somewhat scales down (level-appropriate gear has an edge over synced-down gear, but for most things it’s close enough not to matter). The drawback is that you don’t get skillups while you’re synced down.

  3. Absolutely, well said. There’s a reason why you find cutthroat pirates in Eve Online, the game design and rationale positively encourages it. Fair enough, you’ve targeted an interesting niche for gameplay, great incentivization there.

    A Tale in the Desert fascinates me because of the multiple incentives for both cooperating as well as competing. You observe different folks taking differing paths for success. Long term, cooperation wins out. Short term, competing wins. The differences in individual player philosophy provoke a good deal of debate and conflict.

    From observation, this is how it turns out: the most ‘successful’ players who do both by being social leaders, fostering cooperation and community, as well as being extremely hardcore and progressive on an achievement front to sweep up the competitive aspects first. Followers who cooperate will take their turns to succeed next, it just takes some patience. Those who chase competition at the expense of cooperating with the community get a short burst of victory but fry their social reputations – they last a couple months at best, then promptly quit. It’s like a little microcosm of real life ingroup/outgroup formation.

    Looking forward to see if Guild Wars 2’s design will manage to change up the way MMO players interact with each other to a significant extent, or if the WoW mentality is too well ingrained into the bulk of the population.

  4. Nail on the head, make cost-benefit analysis work in the direction you want the players to go.

    As to groups and loot. I like special named items on special named boss mobs so what i’d do is:

    1. Dead boss.
    2. Everyone in the group can loot.
    3. The loot is one choice from a list like the loot bags in Warhammer public quests: item, runestone, cash.

    Example: warrior, wizzie and healer kill a named minotaur who drops a good warrior axe. All three lean in to loot. The warrior chooses the axe while he wizzie and healer choose the runestone. Next day they go kill a necro for a good caster robe. The caster chooses the robe, the warrior and healer choose runestones. The warrior slots the runestone into his axe. Third day they all go kill a shaman mob for a good healer staff. The healer takes the staff. The other two choose runestones. All three then slot their runestones into their individual items.

    Make all the gear upgradeable with items that only drop off the same named mobs as drop the cool items so everyone has an incentive.

Comments are closed.