Social Engineering in Online Games

I'm still processing Raph Koster's high speed broadband brainblast on user manipulation and management talks involving network theory, emergence, six degrees of separation and power laws. This is stuff I usually hear from webloggers. It felt like a harmonic convergence to hear one of the game industry's thought-developers using these concepts to explain experiments run with large numbers of paying game-players in massively multiplayer online games. Raph described social engineering without flinching, perhaps a concession to the audience of aspiring online game-based social engineers.

It was fast and sweeping; as Katherine's friend from Georgia Tech pointed out: truth through speed. Though he did encourage the development of small, boutique online communities, that's not the business that he is in. It was unsettling seeing his studies of social dynamics harnassed to build effective subscription-based online life simulators for multinational megacorporations.

(These are notes from his lecture, taken in a loose outline using Radio. Most links below are my web searches based on his references. In addition, Raph has put his PowerPoint presentation online.)

Small Worlds: Competitive and Cooperative Structures in Online Worlds
A talk by Raph Koster
at the Game Developer's Conference, San Jose
Thursday 06 March 2003

• Graph Theory - the science of taking a point, and another point, and drawing a line between them
• people are nodes, relationships are lines
• network science - six degrees of separation
• Granovetter - "Strength of Weak Ties" - networking
• You probably got your job through a weak tie
• Watts and Strogatz - "clustering coefficient"
• You want to strengthen weak ties in your online gaming community
• target hubs - market to hubs
• let people join multiple guilds
• multiple interests
• you can only reach 24% of the web by linking.
• How to reach the silent majority of players?
• get a representative sample
• Social Channel Capacity
• Online game developers, on your bookshelf you should have books about: Psychology, Sociology, Anthropology, Economics
• Humans make few social connections - hard-wired to the size of your cerebral cortex
• average ~150
• same in Ultima Online guilds
• so make sure your guilds support at least 150
• Scale-free networks - power law distribution
• Online games are social networks - they don't die - they have to be killed
• you have to take out hubs
• Raph speaks to to Will Wright (in the audience) about web popularity rankings: "Will, if you expose the most popular players in The Sims Online, I might offer them free accounts to lure them to my game"
• Relationship Theory
• Links tend to be homogeneous
• In online games, have a mechanism to allow people to spot people who are just like them - Rich Vogel
• Simulations and UO shows that PvPers will take over environments.
• Unpredictable policing: high-profile, high penalty sweeps are a bigger deterrent to misbehaviour than measured consistency
• act like Gordon Walton - he banned 300-person guilds (guilt by association with just a few people)
• customer service calls dropped through the floor - fear spread of a lightning bolt from god
• The tipping point is good stories
• Tell users of new features, benefits of game, not through bulleted lists but stories
• online game developers suck at this
• Mathematically, if you want an idea to infect a group, target a hub
• any preferential targeting of a message is better than random targeting
• Linda Price of U. Nebraska studies "Mavens" - experts on areas of knowledge
• There is no way to market to the casual market
• you market to the hardcore, and they market to the casual
• Online game designers must be on the message boards at player2player, unknownplayer, waterthread.org
• sadly Lum the Mad is gone - used to be single hub for marketing to online game community
• Bruce tracks subscribers to online multiplayer games
• Game Theory - optimal conclusions from interactions
• 90% of human interaction is done through body language
• William Condon - "Synchrony"
• the single biggest thing you can do to increase relationships in your game - increase the social bandwidth
• Robert Cialdini - six strong persuasive characteristics in humans.
• "broken windows" - eliminating petty crime lowers crime over all
• Best thing to ban in a game? 1337 speak names
• Zimbardo's 1971 Stanford Prison Experiments
• Hartshorne and May - "experiments on cheating"
• honesty is not inherent, it's situational
• Good Samaritan talk experiment
• Fundamental Attribution Error - people underestimate circumstances and overestimate character
• Micronesian suicide epidemic
• We follow our peers - copycatting
• Never publicize behaviour in your game that you don't want imitated
• Only publicize behavior that you want followed
• Prisoner's Dilemma
• optimal response - generous tit for tat
• be nice, be retaliatory, forgive, be transparent
• reciprocal altruism hardwired into the pleasure centers of our brain
• if humans are told "that's a faceless corporation or a computer" in the other room - then they screw them every time
• make your community managers human, every chance you get.
• Online games are a subset of sociology
• Jared Diamond's social complexity of tiers based on population size
• bands, tribes, chiefdoms
• large communities develop stratification and even slavery
• as a newbie in a large EQ guild, you are ordered to mine 500 gold before you can join the hunting party
• so keep a cap on your communities - don't let them go over 500 people
• you get ugly social structures
• mere peer pressure doesn't function
• Social, economic development of online games is retarded by part-time involvement
• find ways for players to be economic participants 24/7
• offline vending, crafting, fees
• Humans are driven by a need for food
• Experience points closest in-game analog to calories
• gamers band together to make that process more efficient (hunting parties in games like agriculture - "farming")
• Lemon problem - Reputation
• Use positive reputation system
• Negative systems blow up
• UO is an example
• Yamagishi, researcher in Japan proved it
• Pareto's Law - 80/20 rule, Power Laws everywhere
• Counter-Strike - zero-sum game - one winner one loser
• fun flows towards the winner
• "funny that the math equation for this phenomenon has the letters 'P' and 'K' in it"
• In a game run by the hardcore players, fun links point inwards towards them
• Google your game - count the number of hits for "upcoming games"
• you will be able to plot yourself on a graph that remains constant - you can only take the place of something
• The curve of growth for every game is fixed early on
• If nodes stop growing after a certain age the entire market is capped in size
• so don't bring games to market that don't grow the market
• For example, you have to really break ground to make entries into the online shooter market
• Go for the boutique online games market
• Tiger Woods, Babe Ruth, Gretzky, Secretariat, Fisher - superhumans, statistically
• The #1 player-killer in UO had 14,000 murders, versus 2,000 for the runner up and 2 for the average
• If you have fixed games, the ladders will develop a fixed top
• the average person is below average in a skill-based endeavor
• handicaps and leagues develop same curve, but at least shallower
• The MUD model of rewarding time playing gives everyone a chance to get to the top
• Conclusions
• online worlds are susceptible to social network analysis
• invite the most popular/respected board posters and give them info and coach them.
• increase the bandwidth of the social interaction
• (Extensive bibliography listed
• Psychology of Human Misjudgment
• 10 Secrets of a Master Networker

Comments

[shadok]

I found the full powerpoint presentation a fascinating read, very dense and wide-ranging (the one on his web site is about 35 slides longer than the one he presented at GDC) and I wish I could have been at the presentation (too bad they didn't webcast it). I don't think it's a successful checklist of Things One Should Do in an online game, but as a grab bag of Things One Should Know And Keep In Mind for anyone designing, implementing or administering a MMOG it should be required reading.

There's some good discussion about the talk on the Waterthread boards where Raph participates. http://www.waterthread.org/forums/

Something he said that jumped out at me: "For me, the best (and most depressing) thing about going to GDC and doing a talk like this is to run down the list of stuff you SHOULD do, and realize how much of our own advice we're punting on from lack of time and or money. Keeps ya honest."

[Timothy Burke]

I've just posted some of my own thoughts on the GDC aftermath, at http://www.swarthmore.edu/SocSci/tburke1/gamefutures.html, riffing mostly off of Greg Costikyan's blog.

[Daniel B. McMillan]

It's the �worm in the apple syndrome.� If too many are bad � and families start to become active in what their kids are doing - we all get framed for being part of the problem rather than noticed for developing a better option.

We know that all crimes begin in the mind. Therefore, in a virtual reality, the mind must be disciplined through the support of positive and constructive behavior. Careless game play � or calling it just a cartoon for the sake of exploiting or making criminal behavior seem rewarding or humorous reinforces destructive thought processes.

In many games, the main involvement revolves around the combat system. This prompts the defensive and aggressive posture in each of us. It gets us thinking about killing. Society calls that �pre-meditated murder� but in virtual reality only characters are being killed. Even so - the character being killed has been developed over time and through the concern of the user. So in effect, it is like killing a little part of the user. This has been proven time, and time again in the reactions players have when the characters they have spent hours developing are killed by "griefers." Thus the MMORPG presents a new equation for game development.

The solution is simple. Foster in-game accountability for one's actions, in order to reinforce the notion of being civilized in the place where it all begins � the mind. This is just a glimpse into the new challenges virtual reality presents to developers as new virtual frontiers become more and more authentic, and the line between reality and virtual reality is perforated. In any case, time will present new studies in the psychology and sociology of the human condition as next generations spend a portion of their life in virtual reality

Freedom is not being able to do whatever comes into your mind. Freedom is being able to choose the right thing.

-Daniel B. McMillan
Creator: Frontier 1859 MMORPG