Monday, 28 February 2011

Review - Human group formation in online guilds and offline gangs

1. Johnson N, Xu C, Zhao Z, et al. Human group formation in online guilds and offline gangs driven by a common team dynamic. Physical Review E. 2009;79(6):1-11. Available at:

This paper looks at building a simple model of how people decide to join or move between guilds or gangs, using membership data from World of Warcraft guilds and LA street gangs.

They attempt two types of model: where people choose to join groups where all members have the same attributes as them ('kinship') and where people choose to join groups where they complement the existing membership ('team formation'). The model that fitted the data most closely was the team formation model. I think that's extremely interesting, as I'm suggesting that guilds do form from players with similar playing styles, so perhaps this is a different angle. (By playing style I think I'm meaning subtle differences in emphasis, i.e. more emphasis on exploration vs raiding vs PvP etc.)

In addition to looking at both sets of data as a whole, they then divide the groups by 'ethnicity'. This is obvious in terms of the gangs, but for the WoW data they split it by server. This turns out to be a good fit because as it is not possible to change your ethnicity in real life, it is also difficult/expensive to move your character between servers in WoW. They find that although the model fits all the variations of ethnicity they need to tweak some of the constants slightly, suggesting that there are very subtle differences in behaviour not only between different gang ethnicities in LA (which might be expected - different backgrounds and cultural values) but also in between servers in WoW (same game, so what causes that?!).

Good analogy to job finding, and actually rather tallies with my empirical experience that teams tend to take on the character of their head (who does the hiring) while looking for members that complement the group skillset. Again, a difference between the two definitions of what people look for in a team - similar values, different skills.

Thursday, 24 February 2011

Tragedies of the ludic commons - Jonas Heide Smith

1. Smith JH. Tragedies of the ludic commons - understanding cooperation in multiplayer games. Game Studies. 2007;7(1). Available at:

An article that links player behaviour and dilemmas in online games into two new areas of research for me - collective action problems and social dilemmas. This is sort of what I've been looking for, because I think that research on multiplayer online society ought to have some grounding in research into offline societies, even if it's just to expose the differences between on and offline experiences.

The paper introduces the concepts of collective action and social dilemma, but then links the two for the remainder of the paper. He then discusses three aspects of multi-player gaming that he describes as 'conflict-heavy' and how they can be understood in terms of social dilemmas. The three areas he chooses are aligned with my interests: cheating, grief play and responsible participation.

Cheating he acknowledges is a slightly grey area in places. Defining some cheats is straightforward - hacking the client for example obviously contravenes the rules of the game. However, exploiting loopholes that the developer didn't expect is grey - the player has to guess at what is beyond allowed, and what the developer expected. Also the author states that if everyone is 'cheating' it may just be that they are all playing a different game (although mostly the games where everyone cheats are not wildly popular). The cheat is compared to the prisoner's dilemma, where the very best outcome for the player is to have everyone around them be playing fair while they cheat, but the best for the group as a whole (in terms of the game-playing experience) is if everyone is playing fair.

Grief play is defined as disruptive play that brings the player no or little personal gain. If the collective good is again seen to be the enjoyment of the experience, this can be seen to destroy that collective good.

Irresponsible participation is a tricky one. This is defined (in this article) by example - so if you know a raid is going to take upwards of an hour and you only have 40 minutes, you'll have to abandon and potentially wreck the raid party. Alternatively, gambling at one point in a game and putting yourself at such a disadvantage that you quit could take the enjoyment of beating you from your opponent. Difficult. I suspect these might be very personal calls on what is and isn't acceptable - would be interesting to follow up on that.

The article goes on to discuss the ways that these problems could be alleviated for the good of all, including code, cheat-blocker software etc. It gets most interesting for me when he starts to talk about guilds as a method of managing the great proportion of these issues. The guild rules and reputation allow the players to judge others - so you know that if you're playing with people from your guild they ought to have the same ideas about what is and isn't permissible (within the norms you are forming). So, if a guild is one where team kill is turned on, you know that and understand that risk and won't be so upset if one of your team shoots you (in theory). I think the grey areas and judgement calls (i.e. is it acceptable to concede the game early and deny the opponent the thrill of winning) is what the guild rules will probably centre on and clarify - it would be fascinating to check that.

Tuesday, 22 February 2011

The Media Equation, Byron Reeves & Clifford Nass

The media equation (published 1996) states that “media equals real life” and is, the authors assert, a rule that governs how all of us interact with media. I guess most of us can imagine equating mediated and real life as something that we do occasionally when tired or stressed - shouting at the computer, for example - but the authors insist that it is ubiquitous - all of us do it, and do it often. To test their hypothesis that our interactions with computers, TV etc. are fundamentally social and natural, the authors started with social science research into human behaviour. They chose a range of findings, substituted “person” or “environment” with media to form a new hypothesis which was then tested in the same way as the original social science finding. For example, in investigating politeness, the authors came up with the following hypotheses:
  1. When a computer asks a user about itself, the user will give more positive responses than when a different computer asks the same questions.
  2. Because people are less honest when a computer asks about itself, the answers will be more homogenous than when a different computer asks the same questions.

Areas they investigated included manners (politeness, interpersonal distance, flattery, judgement), personality, emotion, social roles (including gender). In every case the authors found that their research backed the hypotheses.

They also found that it is true for all sorts of media – computers, interfaces, pictures, audio, video; it is true for all types of people (whatever their education or background – computer scientists included) and it isn’t dependent on the complexity, sophistication or “realism” of the media – very basic images or interfaces will elicit the same responses.

The authors are quick to point out that it’s not that we think computers are people, or that we think it reasonable to act as if they are – in a sense we can’t help doing it. (well, we can avoid it for a time but it requires a lot of effort and is difficult to sustain). And the reason behind these behaviours? According to the authors, evolution is the culprit – it simply hasn’t equipped us to deal with technology. Up until the technological era only humans exhibited rich social behaviours and all perceived objects were real objects, so we have evolved to respond to anything that appears to behave socially or seems to be a real object as if it actually is.

These findings have obvious implications for the design and evalution of media. If we design systems that mirror how social relationships and physical environments work, users will find using these systems intuitive, will more readily engage with the systems and find interaction more rewarding.

Wednesday, 9 February 2011

Jane McGonigal

Has done a rather interesting TED talk. I suspect it briefly touches on some of the subjects of her new book: Reality is Broken.

She claims she wants people to spend more time playing online games to solve the world's problems. This is based on a few things, including the idea of the emotion of the 'epic win'. An epic win is an outcome so positive that you had no idea it was even possible to achieve it until you are on the verge of it. These situations happen in games, but rarely in real life. She says the more normal feeling in real life is 'I'm not good at life' - or the feeling that you are not as good in real life as you are in the games.

She goes on to look at why games are making us feel so good about ourselves. Game challenges are always available, always sound epic or important and yet are carefully targeted to the ability of the gamer so that they are on the edge of what they can achieve. No unemployment, no twiddling thumbs or time to wallow. In addition to that you get plenty of positive feedback in terms of levelling up, which is missing in real life.

There is an incredible statistic that people have so far spent 5.93 million years online playing World of Warcraft. I think that must be in man-years. That's similar to the amount of time taken for humans to evolve from just standing. Apparently youngsters in countries with significant gaming cultures will spend 10,000 hours playing online games by the time they are 21. Apparently around 10,000 hours is the time taken to become a virtuoso, so she goes on to look at what gamers are learning from playing those games:

  1. Urgent Optimism. There's something that needs doing immediately, and it Can Be Done. 
  2. Weaving a Social Fabric. Proven to like people better after playing games with them (even if they win) because you've trusted them a lot to play the game in the first place. I think there might be some shared experience in that too (see cricket team!) but need to look into references. 
  3. Blissful Productivity. Humans are happiest working, not relaxing. Gamers will work for hours if given the 'right work'. 
  4. Epic Meaning. Gamers weave truly epic stories, and are used to operating with tasks that have a big effect. 
The problem is that they are changing virtual worlds rather than real worlds. The economist Edward Castronova describes what's happening as a mass-exodus from real to virtual worlds. Jane McGonigal believes that people are escaping their broken real world problems to virtual worlds, where they can form better social connections, and feel like they are good at it. She's working on games that bring those powers of concentration to the real world. Three so far, and I'd like to follow up on all of them:
  1. World without oil - piloted in 2007. 
  2. Superstruct
  3. Evoke - played in 2010.

Evoke looks particularly relevant as it was shared with universities across sub-Saharan Africa.

Thursday, 3 February 2011

Review - Homo Ludens, J. Huizinga

1. Huizinga J. Homo Ludens: A study of the play element in culture. Beacon Press, Boston; 1955.

A lot of the books and papers I'm reading kept referencing this book, so I thought I ought to read it. It's been a bit of a struggle, frankly! The most interesting sections for me included chapter 1: Nature and significance of play, chapter 3: Play and contest as civilizing functions, and chapter 12: Play-element in contemporary civilization.

Chapter 1 includes good definitions of play, that are built on by Jesper Juul and others. Play is "non-seriousness" and is outside of "normal life". It takes place in a magic circle, with boundaries carefully built up and maintained in the game by the rules. Huizinga states that the rules are "absolutely binding and allow no doubt" (pg 11) - I'm not sure this holds for multiplayer (or maybe computer in general) games where the rules of the game are discovered as the players play, although it could be argued that the rules of the game as defined by the code are in place before play starts, even if the players don't know them all yet. If the rules are broken or have to be renegotiated the game has to stop for that to be fixed.

In addition to this he introduces the concept of the play-community, the group of players, which can become permanent after the game ends. Along with this is the idea of the 'spoil-sport' who refuses to play the rules any more, and therefore exposes the players to the fact that their game is just a game. The cheat is something different, who keeps the magic of the game going by pretending to play but secretly breaks the rules.

In chapter 3 there is some discussion about single player vs multiplayer. Social play (or multiplayer) is considered more "generative of culture" than solitary play. It's recognised that the prevalence of single player games is quite a recent phenomenon (not in this book, in more recent literature) that came about with the personal computer.

(There's a cracking quote too: "The passion to win sometimes threatens to obliterate the levity proper to a game." pg. 47 - could be relevant in terms of why people might be driven to cheat?)

Then towards the end of chapter 12 the following: "Civilization will, in a sense, always be played according to certain rules, and true civilization will always demand fair play. Fair play is nothing less than good faith expressed in play terms." I felt the chapter on the play-element in contemporary civilisation was rather reflective of it's time shortly after WWII, but this particular point could again relate to a nebulous concept of 'fair play' and how that is defined by the community.

Relatively few points for a book I really struggled with!