Social Mechanics

8/8/2019 Social Mechanics

1/162

Social MechanicsThe Engines BehindEverything Multiplayer

Raph Koster

VP Creative Design


2/162

GDCO 2010: Raph Koster, Social Mechanics: the Engines behind Everything Multiplayer 2

What are we going to do?

Go very fast.

Over 160 slides.I could not restrain

myself.


3/162



The core 40mechanics ofmultiplayer

games. Yes, all of them.

No furtherdesign required-- ever.*


4/162



The core 40mechanics ofmultiplayer

games. Yes, all of them.

No furtherdesign required-- ever.*

* For certain values of ever.


5/162



Theyll be markedlike this:

For your tweetingconvenience!


6/162



We will touch briefly ongame grammar


7/162GDCO 2010: Raph Koster, Social Mechanics: the Engines behind Everything Multiplayer 7


and on economictheory




The famedpsychologist Stanley

Milgram




1940s science fiction




a game called Werewolf

And




a somewhat lengthy

discursion into

structuralistanthropology,

for which I apologize, butnot as much as for thetime spent on




a little

Derrida.



But first

A good game should focus entirely on it'ssingle-player aspect first and foremost. Then ifit's a simple game like a shooter or racer, use the

remaining time and space to fit in a multi-playeraspect to it In other words, multi-playing

should never take away from the single-playeraspect of the game. Pure multi-player games

really should be few and far between.

-- the unfortunately named New Troll, Escapist Forums, 23 June 2009


15/162

True single-player

1 v. self/system



Defining a single player game

A game that is not played in opposition to, orin parallel with, or collaboratively with,

someone else.





someone else.





someone else.

(In other words, hardly anything.)





someone else.

(Especially since you can regard the

computer and physics and your ownbody as being your opponent)


21/162


Boundaries

The boundary of a game is not the board,event, world, etc.


22/162


Boundaries

Consider: is training part of the game?


23/162


Boundaries

How about doping?


24/162


Boundaries

The true boundary of a game is based onwhere actions that can legally affectgameplay stop. Training yes, doping no.


25/162


Helping

The simplest form of multiplayer, presenteven in single-player games, is simple

advice and assistance.

An outgrowth of spectating.


26/162

Parallel symmetric

1 v system vs 1 v system


27/162


Parallel symmetric games


28/162


Status

Quantifying a players achievement againstthe systemic opponent.

Typically becomes multiplayer when relativestatus is compared.


29/162


Races

The first user to reach a given status level wins.(The basic boardgame, and basic sport)


30/162


Leaderboards

Compete in asynchronous parallel with allhistorical attempts.


31/162

Parallel asymmetric

1 vs system, vs 1 vs another system


32/162


A hypothetical arcade game

The player is playing nested games of1. Steering the ship

2. Hitting the asteroids with lasers

3. Dodging asteroids and bullets and UFOs

4. Getting advice from her friend over the shoulder

5. Getting as high a score as she can

6. To brag to her friend that she beat him

7. And competing against everyone who has played thisparticular arcade machine

This nesting quality will be important


33/162


Scaling parallel

Scaling parallel symmetric play to largenumbers is hard to do in a time-boundedway.

For example, a few people can play an arcadegame one after the other and compare

or an infinite amount can play againsteveryone ever via the high score table

But how do you do a medium size number?


34/162


TournamentsBrackets are often a

game of their owndistinct from thescoring or status

comparison.

Of course,

tournaments areoften used for


35/162

Opposition

1 vs 1


36/162


Rival and non-rival goods

A rival good is something thatcannot be used by someone else

at the same time.

A non-rival good is stuff like

information, or the publicpark.


37/162


Flower-picking

Non-zero-sum resource consumption: Youcompete to get stuff, but there is alwaysmore stuff.

Speed-up arrows on a race track even pickups in many games.


38/162


Dot-eatingZero-sum resource consumption.

(also territory, in a graph theory sense).

Players compete to get stuff, and whatever I get,you cant have.


39/162


Tug of war

I can take your stuff, you can take my stuff.

Exactly like zero-sum status.

The engine behind all combat games.


40/162


Handicapping

Artificially equalizing status inorder to provide a tighter race.


41/162


Secrets

In game theory terms, imperfectinformation knowledge as a rivalrousgood.

Fog of war, hands of cards, etc.


42/162

Multifront opposition

1 vs 1 vs 1 vs n


43/162


Gutscheras terms

At GDC 2009 K. Robert Gutschera used differentterms than mine:

Orthogames: games with a statistical ratingoutcome

Parallel symmetric games: races (OK, same termhere)

Brawls: gluing together 2 player orthogames.

Politics start when you have a choice of whom tointeract with.


44/162


Rival goods mechanics

Several forms of brawls depend on treatingresources within the game, such as

Health

as rival goods.


45/162




HealthSurvival

as rival goods.


46/162




HealthSurvival

Stuff

as rival goods.


47/162


Take many tugs of war


48/162


Last man standing


49/162


50/162


Social effects depend on humanization

Numerous social science experiments havefound that a variety of psychological effectsonly occur when we think the interlocutoris human, or like us.

The Milgram buzzer experiment

Zimbardos work

(Mirror neurons are all the rage.)


51/162


Booing

The multiplayer evolution of helpingA psychological 3rd party attack it only

works on human opponents.

d bl ff


52/162


Once secrets are treated as goods,misinformation becomes a form of tug-of-war.

Deception and bluffing

d bl ff


53/162


Once secrets are treated as goods,misinformation becomes a form of tug-of-war.

Deception and bluffing

d b


54/162


3rd party betting

3rd

parties play a bidding game against arivalrous resource pool, based on theoutcome of a multiplayer game.

P i Dil


55/162


Prisoners Dilemma

Teammates with secrets from each other whomust choose independently to collaborateor both lose.


56/162

K i i l


57/162


Kriegspiel

Credited to Von Reiswitz First used around 1812 as a form of training for

the Prussian army

K i i l


58/162


Kriegspiel

Popularized by science fiction writers H G Wells wrote Floor Games (1911) and Little Wars (1913)

Which led to Fletcher Pratt designing naval wargames in the1940s

K i i l


59/162


Kriegspiel

Some of whom went on to invent the thinktank.

K i i l


60/162


Kriegspiel

Some of whom went on to invent the thinktank.

G


61/162


Gamesmaster pattern

A multiplayer gamethat relies on a thirdparty to direct thegame an evolutionof the non-mechanicreferee.

Leads directly

to

(Us in

thisroom)


62/162

Small groupsn vs n

W lf


63/162


Werewolf

Aka Mafia, etcA game where players have secret roles

Villager, werewolf, various other small optional

ones

W lf


64/162


Werewolf



ones

W lf


65/162


Werewolf



ones

W lf


66/162


Werewolf

Wolves get to kill off villagers, Villagers getto lynch anyone

Tons of these mechanics are present:deception, Prisoners Dilemma, brawls,

gamemaster

W lf


67/162


Werewolf

Wolves get to kill off villagers, Villagers getto lynch anyone

Tons of these mechanics are present:deception, Prisoners Dilemma, brawls,

gamemaster

R l


68/162


RolesSpecialization into different games within one

group

H t t t / i


69/162


Hot potato/ganging up

Rotation of roles within a multiplayer game(often a brawl), creating ad-hoc groups.

Think of tag



70/162



Rotation of roles within a multiplayer game(often a brawl), creating ad-hoc groups.

Think of tag

Tag


71/162


Tag

Layering of multiplayer mechanics gives way towhole new games

Add secrets: Blind Mans Bluff.

Add dot eating: Duck Duck Goose (compete for the

chair) Addflower-picking: Freeze tag.

Addgrouping: Cops and Robbers.

But then theres Olly Olly Oxen Free! (orwhatever) whats that about?

Rituals


72/162


Rituals

Ceremonies marking significant roletransitions within a social structure.

Functionalism


73/162


Functionalism

Sociology and anthropology used to see all ofgroup interaction as being about structure:groups as organs that make the bodywork.

Today the field has moved on, but its auseful lens for multiplayer game designers!

Ritualism and lifecycles


74/162


Ritualism and lifecyclesCompare Robert Mertons theory of deviance to

player lifecycles:1. First, users try to conform to the rules as they work to

understand them.

2. Then they try to innovate and reach the goals in newways.

3. Then they keep doing things the right way but stopcaring about the objective. This is called ritualism,and he has a great Hopi rain dance example.

4. Then theyretreat and stop caring about the goal or

the method.5. Finally, theyrebeland start doing their own thing.



75/162




understand them.







76/162




understand them.







77/162




understand them.







78/162




understand them.







79/162




understand them.






80/162

Common rituals


81/162


Common rituals Transitions of state (birth, marriage, death,

levelling up) Holidays (environmental transitions,

commemoration of historic or individualtransitions of state)

Religious (requests for intervention)

These are often marked by mechanics such as Changes of uniform

Gatherings But above allgifts.

Common rituals


82/162





Religious (requests for intervention, comfort)



Common rituals


83/162





Religious (requests for intervention, comfort)



Gifts are ubiquitous


84/162



We give away the bride



85/162




We give gifts at major holidays



86/162





We gift the land with a bride at solstice



87/162






We give gifts to the dead to take with them



88/162







We give gifts at weddings



89/162








We give gifts to babies



90/162








We give gifts to babiesWe throw t-shirts at hockey games

Gifts


91/162


Gifts

Transferring a rivalrous good to anotheractor in order to increase their status.

Gift theory


92/162


Gift theory

Marcel Mauss in Essai sur le Don Gifts are never free to either party.

Reciprocal exchange is typical. Failure to return a gift in Polynesia costs you mana.

There are three steps: give, receive, reciprocate.Each can fail and each can have consequences.

Gifts are inalienable they always retain theirsource.

Kashmiri anecdote


93/162


Kashmiri anecdote

A Kashmiri tale tells of two Brahminwomen who tried to fulfill theirobligations for alms-giving simply bygiving alms back and forth to oneanother. On their deaths they were

transformed into two poisoned wellsfrom which no one could drink,reflecting the barrenness of this

weak simulacrum of giving.

-- Wikipedia, summarizing an anecdote itself summarized inLewis Hydes The Gift: Imagination and the Erotic Life of

Property,who got it from who knows where

Reciprocity


94/162


Reciprocity

In an iterativesimulation, thesocialexpectation of

a future returnof a gift.

Tiers of economy


95/162


Tiers of economy

Non-market economies: no concept of price

Generalized reciprocity:just give, whee, no expectations! Reciprocity resides in satisfaction, social closeness. In western

society this tends to be gifting between family members, etc

Balanced reciprocity: expectation of a return gift at an

undefined future date. Still pretty informal. This is where social game gifting tends to

reside.

Negative reciprocity: barter systems, which enable

transactions between strangers. A way to establish friendly relations when meeting strangers.

Mentoring & twinking


96/162


Mentoring & twinking

Onboarding and socializing group members via the

gifting of rivalrous or non-rivalrous goods. Giving a car at graduation

Or Mickey Mouse ears

when you join Disney.

An asymmetric

exchange: I getsocial obligation,

you get value.


97/162

Identity


98/162


Identity

Means of displaying status and role viarivalrous goods.

Extremely important with groups, from skinsvs shirts to class gear, etc.

Ostracism


99/162


Ostracism

Group removal via denial of commonresources.


100/162

NetworksScale-free social structures

Clustering


101/162


g

Until now we have talked about groups,which kind of look like this:

In social networking theory, these are called

clusters. And they often have weak ties to other clusters.

Clustering


102/162


g




Clustering


103/162


g




Networks


104/162


Technical term: scale-free networks.

A very particular sort of structural pattern and universal in human society.

Preferential attachment


105/162


Rich get richer syndrome A new person coming into a network tends to

pick the most popular node to connect to.

Power laws


106/162


The mathematical distribution of popularity of

link destinations.

There are consequences to this distribution. Median lower than the mean Unknowability of the network Unequal information dissemination Network longevity and hub vulnerability

Bose-Einstein condensate


107/162


Aphase transitionwhere a networks powerlaw gets so extremely kinked that a singlehub dominates:

Windows operating system

Tiger Woods win-loss record earlier in hiscareer

The winning kingdom in Shadowbane

Iterative simulations


108/162


Say we all roll dice against one another over and over

But every time I win, I get to add +1 to my future rolls.

A power law happens fast.

Lets say instead that some users just have a +3.A power law also happens fast.

Competitive arenas with any sort of persistence lead to the

average player being a loser.


109/162



110/162










111/162








Channel capacity & sympathygroups


112/162


groups

Lots of evidence to suggest thatyour knowable part of thenetwork is limited Robin Dunbar, monkeysphere

etc

This leads to special mechanicswhen dealing with nodes atdifferent graph distances.

Trust


113/162


Trust up close and its non-transitive!

Reputation at midrangeFaith at a distance (aka social contract)

Homogeneity


114/162


g y

Tightly linked nodes tend to be homogeneous liking the same things

This also introduces interesting liabilities ingameplay

As those with kids know, ones peer groupcarries more social weight than authority

Combined with trust and other networkcharacteristics, we get


115/162


116/162

Guilds vs Neighbor Structures


117/162


T. B. Naik specifies these as the identifiers oftribes:

social distance from other tribes

frequency of contact

community of interests

scale compactness and self-sufficiency

politically organized

customary laws

multifarious economic pursuits

Impact of scale


118/162


Different social structures at different scales:

Bands

No fixed home"Egalitarian"leadership

No real bureaucracy

No laws

Unstratified culture

Tribes

A single home"Egalitarian" or "big-

man"

Organized resourceextraction

Still unstratified

Chiefdoms

Many homesCastes and classes

Cronyism & monarchs

Bureaucracy & laws

Taxes, indentured labor,and slavery

Public architectureLuxuries for elites

Dozens Hundreds Thousands

Segregation and genocide


119/162


Mathematical simulations show that given apetri dish, tribal structures always

Self-segregate

Annihilate each other

Exclusivity


120/162


The use of identityand other earliermechanics tomanipulate thevalue of rivalrousand non-rivalrousgoods.

Also known as velvetrope.


121/162

Massive groupsDesigning games at the societal level

Swapping models


122/162


The functionalist way to look at things is:

Game systems that are problems of control

and coordination at a societal level-- Herbert Spencer

Resource distribution, goods production, etc.

Structuralism


123/162


The sequel to functionalism! More of anemotional model, also now out of favor:

Meaning produced and reproduced viapractices and activities that are systems of

signification-- Levi-Strauss

(Such as games)

Guild vs guild


124/162


Large-scale interactions using availablesystems to create competition betweensocial entities

Trade and contract


125/162


Formalized means of allowing mutuallybeneficial interactions between overly-distant nodes on the network

MMOs versus social games


126/162


MMOs developed stronger social structuresthan social games, because they had sharedeconomic participation

despite not having constancy.

Social games usually have constancy withoutshared economic participation!

Anthony Giddens


127/162


Social structure as game rules

Ant farming or goldfish bowl

Structure, Modality, Interaction

Structure is the rules and resources

Modalityhow structures become action.

Interactionwhat the actor actually does

This analogizes directly to the forms of rulesin Rules of Play, or to the MDA framework.

Post-structural thinking


128/162


A challenge:

A lot of post-structural thinking is aboutseeing past dualities and to shades of gray.

Moving beyond dualities such as neighborand not, tribe or not.

Games do this poorly. But

The PASSE model


129/162


Five general ways of building large, shades-of-gray social

structures in multiplayer games:

Politics

Arts

Science

Sociology

Economics


130/162

PoliticsInfluence and social control

Elections


131/162


Actors of any distance from the

interaction apply a rivalrous good towardsan objective metric that applies

network-wide.

Politics --

orAmerican Idol,the largest MMO

in the world.

Reputation, influence and fame


132/162


Quantified signifiers for social currency, andformalized spending of this rivalrous good.

Conflict theory


133/162


The European post-structuralism:

Structures are developed out of conflictbetween people with differing interests andresources.

Bartle Types


134/162


Structure vs agency


135/162


The big question: are we locked into thegame design?

Structure creates habitus a structure of mind,a set of acquired schemata what goes

without saying for a group -- Pierre Bourdieu

The same question as socialization vs

autonomy.

Structuration


136/162


Anthony Giddens in 1984 had an answer:

Social life is both the small andthe big

People constantly reinforce and extend thestructure through actions

And the structure constrains what they can do

So theres a reflexive action that makes things

change overall, slowly.Which leads to interesting game problems such as

Public goods


137/162


Like a multiple player prisoners dilemma.Public goods are non-rivalrous, so there is atemptation to enjoy the good with nocontribution. For example, breathing.

Public parks.

Tragedy of the Commons


138/162


Flip side of public goods: if they are rivalrous,then they can get used up. Classic example,the commons in a village. Or spawns,territory, etc. The game is in

solving the

unsolvable.

And finally

Community


139/162


Where we start playing games on YOU.

A properly managed community isstructuration in action, and becomes the

ultimate political game for users.


140/162

ScienceFiguring out how things work

ARGs and Puzzles


141/162


Ridiculously complex puzzles can be solved by hive

brains.

ARG puzzles such as image-based steganography

hidden in white noise in a poster. Combat systemalgorithms in an MMO.

Underlying logical rules discovered viaexperimental method equals

Strategy guidesh f f l b d


142/162


The transformation of rule-based secrets into

a public good.


143/162

SociologyThe guts of sociability

MLS


144/162


Multilevel Selection Theory

Natural selection can be affected bycooperation at the group level.

-- David Sloan Wilson and Elliot Sober


145/162

Interdependence and scaling issues


146/162


Scale-free networks are highly insulated from

failure.

But collapse occurs if hubs are suddenly removed.

If your game design has hubs and one breaks

(cfStar Wars Galaxies, Pirates of the Burning Sea)


147/162

Economics

Fundamental multiplayer dilemma


148/162


The game equivalent of psychologys Fundamental

Attribution Error.

The basic premise of economics is mutualimprovement and optimal distribution of

resources.

Its a way for a group to level up.

This is anathema to games.

Veblen goods


149/162


A good where peoples preference for buyingthem increases because of their price.

Luxury cars

Limited Edition virtual items

See also: soulbinding, level limits, etc.

Services


150/162


Classes and roles can be seen as forms ofasymmetric trade.

Even mechanic #1, helping is a service!

So are numerous others on the list.

In other cases, mechanics are premised inunequal contributions.

All of these aregenerally non-quantifiable.

Therefore, cheating


151/162


Randy Farmers KidTrade proposal Gifting Twinking

Gifting + Multiple Chars/ServerMuling

Gifting + Messaging + Trust Trading

Trading Messaging Trust + In World Machinery Robust

Trading Robust Trading + Scarcity+ Liquidity External Market (eBay)

External Market Trust + In World Machinery GOM

Gifts inevitably cause Real Money Trade,which is a form of

All i l tifi d l ti f

Arbitrage


152/162


Allowing unequal quantified valuations of

the same good to arise within the system,and permitting trade that exploits them.

Supply chains


153/162


Cascading unequal valuations ofgoods arranged in sequence.


154/162

Arts

The act of invention


155/162


Games Are

solving statistically varied challenge situationspresented by an opponent who may or may not

be algorithmic within a framework that is a

defined systemic model.-- me

Statistical variations are handcrafted,randomized or

User generated content


156/162


Forums. Character art.

Roleplay narratives.

Fan fiction.

Player conventions. Weddings.

Game maps.

The game has to be designed to encompass and permit thissort of activity.


157/162

DeconstructionTaking it all apart

Remember the quote?


158/162


A good game should focus entirely on it'ssingle-player aspect first and foremost. Then ifit's a simple game like a shooter or racer, use theremaining time and space to fit in a multi-player

aspect to it In other words, multi-playingshould never take away from the single-playeraspect of the game. Pure multi-player games

really should be few and far between.

-- the unfortunately named New Troll, Escapist Forums, 23 June 2009

Derrida

A i d h l i ibl f h


159/162


As a turning toward the presence, lost or impossible, of the

absent origin, this structuralist thematic of brokenimmediateness is thus the sad, negative, nostalgic, guilty,Rousseauist facet of the thinking of freeplay of which theNietzschean affirmation-the joyous affirmation of thefreeplay of the world and without truth, without origin,offered to an active interpretation-would be the other

side. This affirmation then determines the non-centerotherwise than as loss of the center.And it plays the gamewithout security. For there is a sure freeplay: that which islimited to the substitution of given and existing, present,pieces. In absolute chance, affirmation also surrenders

itself togenetic indetermination, to the seminaladventureof the trace.-- Structure, Sign, and Play in the Human Sciences

Griefing


160/162


Griefing can be an act of game literacy:engaging in the ultimate transgressive play

with the system,Or the act of virtual sociopathy.

All* the multiplayer mechanics1 v self/system Deconstruction1 v 1 v 1 v Networks


161/162


f/ yHelping

1 v 1 parallelStatus

Races

Leaderboards

Tournaments

1v1 opposedFlower-picking

Dot-eating

Tug of War

Handicapping

Secrets

Griefing

PASSE:Politics

Art

Science

Sociology

Economics

Last man standing

BiddingDeception and bluffing

3rd party Betting

Prisoners Dilemma

Gamesmaster

n v n (groups)RolesHot potatoRitualsGiftsReciprocityMentoring & TwinkingIdentity

Ostracism

*41-50 are left an exercise forthe reader

Iterative interaction &trust

Guilds

Exclusivity

Guild vs guild

Trade and contract

Elections

Influence and fame

Public goodsTragedy of the Commons

Community

Strategy guides

Teamwork (MLS)

Arbitrage

Supply chainsUser generated content

For brawls and races, see Characteristics of Multiplayer Games: Lessons from the World of PaperGaming K Robert Gutschera GDC2009

Hidden info


162/162

Gaming, K. Robert Gutschera, GDC2009.

The MDA framework is described inMDA: A Formal Approach to Game Design and Game Researchby Hunicke, LeBlanc, and Zubek.

I recommend Linkedby Albert-Lazslo Barabasi for an intro to social networks, and Influence byRobert Cialdini for more on influence structures. And of course you read Rules of Play andDesigning Virtual Worlds right?

For more on network structures, I refer you to my GDC2003 presentation Small Worlds: Competitiveand Cooperative Structures in Online Worlds and its extensive bibliography. Theres a series OnTrust on my blog as well.

I would start with Wikipedia for reading up on sociology and anthropology.

You can read Randy Farmers KidTrade: A Design for an eBay-resistant Virtual Economy on hiswebsite at http://habitatchronicles.com/Habitat/KidTrade.pdf

More on interdependences issues at http://www.next-gen.biz/news/what-went-wrong-with-pirates-mmo

You can read Derridas Structure, Sign and Play at http://hydra.humanities.uci.edu/derrida/sign-play.html or in Writing and Difference. If you really, really, really want to.

All the art came from http://icons.mysitemyway.com in the royalty-free Rounded Glossy Black Icons set.

Special Thanks to John Donham & Jenny Kropko and Patrick Ferland (for PASSE)

This presentation and dozens more are available in various formats on my website:
http://habitatchronicles.com/Habitat/KidTrade.pdfhttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://hydra.humanities.uci.edu/derrida/sign-play.htmlhttp://hydra.humanities.uci.edu/derrida/sign-play.htmlhttp://icons.mysitemyway.com/http://icons.mysitemyway.com/http://hydra.humanities.uci.edu/derrida/sign-play.htmlhttp://hydra.humanities.uci.edu/derrida/sign-play.htmlhttp://hydra.humanities.uci.edu/derrida/sign-play.htmlhttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://www.next-gen.biz/news/what-went-wrong-with-pirates-mmohttp://habitatchronicles.com/Habitat/KidTrade.pdf

Social Mechanics

Documents

Transcript of Social Mechanics