GDC2008: Game Studies Download 3.0

 Posted by (Visited 14717 times)  Game talk  Tagged with: ,
Feb 202008

Mia Consalvo, Ohio University
Ian Bogost, Georgia Tech, also runs Persuasive Games
Jane McGonigal, researcher at Institute for the Future, and makes ARGs

3rd Annual Game Studies Download

We will fill your brains with the ten most interesting research findings of the year. The main question is what do we know about games and game players in 2008 that we didn’t know in 2007.

Game researchers are very smart people who care a lot about games, the people who play them, and the future of the medium.

Targeted expertise in HCI, economics, architecture, and more…

Looked at 100s of articles and studies, then did a rapid fire top ten list with ten ideas and ten practical takeaways, You can downlaod the slides at

#10: The best content understands exactly how the player likes to play and then makes it slightly harder.

Three scientists at U Essex had looekd at ways to procedurally generate tracks in racing games. And they wondered how to make them interesting, especially for different sorts of players. They called this “tuning for entertainment.”

They quickly disocvered that optimizing track performance is not interesting. A fun track is one where the player almost loses control. Drive fast on straightaways but brake fast on turns.

Rather than guess, they looked at how players performed as they raced and used it as a source to generate new kinds of tracks either live or in advance. (shows images of tracks)

Takeaway: custom procedural variations in limited environments can be more fun than big environments and open worlds. How can your next game use player-inspired procedural variation?

#9: Breaking the immersive spell can make gameplay more engaging.

Buckinghamshire Chiltern’s University in the UK. Does immersion really have to be seamless?

Breaking the 4th wall can be good. Clumsy controls can heighten fear and frustration, driving emotional reactions. Games that comment on themselves provide memorabl content and atmosphere. Disruptions can contribute to a unique game style.

He was studying horror games like Silent Hill…

Takeaway: making players remember it’s a game can actually heighten their experience.

#8: reality-based gaming is already a lot bigger than you think.

The first year we did this we were asked to exclude reality-based and location-based games. But now we can look at them because of the market size. Tsinghau University Beijing: are game players getting tired of screens? Very relevant to Chinese culture — a lot of backlash against games, addicting and so on. So there has been a new kind of reality-based game growing in popularity there, that is seen as an antidote.

For example Mafia/Werewofl/Witch Hunt. They developed a gaming console where the game can moderate itself, audio commands, etc, and people play them in real life settings. 1.5m subscribers to go to these location-based clubs with these games. Average age is 20-40, 1/2 are women.

Also developing an array of PDA-supported and laoto-supported games. Majoy City Area game, for exmaple — Counterstrike in the real world. It grew so popular they had to grow it beyond shooters and to dating and shopping and puzzle games. 495,000 players in 7 cities. A non-trivial market that is interested to watch.

Big insights include that in China these reality-based games are considered a better way to play with friends and family. Having standardized tools for fair gaming instead of a human moderator has produced the rapid interest. And that it’s mobile deivces and laptops that enable it.

Takeaway: reality gaming is taking games and mobile devices in a new direction. and off of screens.

#7: gamers can be altruistic, empathic and nurturing, and not just while doing something else but while playing!

Paper at Nordic Games Conference. u Tampere, Finland. Psychological emotion theory and media studies approach to catalog the emotions games achieve.

So can games tap into nice emotions as well as the not-so-nice ones?

In emotion theory there are four sorts to explore:

prospect-based emotions: tied to an event.
Attribution: related tot he emotions of others
fortunes of others: empathy and altruism
Attraction: likes and dislikes

Prospect-based; Zuma… sense of satisfaction. Narrative events in Final Fantasy.

Attribution: emotions of other characters in the game world. Ico, “save the princess” sorts of thing, the whole premise of the game is that you want to remedy the emotion the victim is feeling.

Fortunes of others is where it gets interesting. eliciting sympathy and altruism towards other characters. Animal Crossing, for example, sending gifts. Portal and the companion cube may be the same thing… the object helps us develop a sense of sympathy in the same way as the characters in Animal Crossing.

Attraction: the idea that the aesthetics of the environment cause emotions, like Katamari or unfriendly, like in Silent Hill…

Most games we see privilege the first two, that’s what we do well and pursue most frequently. There’s unexplored design space in the other two, especially in the fotunes of others.

Takeaway: empathy, altruism, and so on can benefit ANy sort of game. How do you make you rplayers want to be nice to the characters in the game?

#6: it takes ten hours of gameplay for women to play witht he same spatial attention skill as men.

U Toronto, part of an NSF grant, how do we get more women into science, math, etc. Videogames were part of their experimentation.

They did a first experiment where they looked at field of view tasks, etc. Gamers vs nongamers, and they found gamers were better than nongamers. Then they recruited men and women to play videogamnes. Control group used 3d puzzle game, and the other group used Medal of Honor Pacific assault, and they tested them at the start and end.

They were looking at spatialattention, what you bring to the process of rotating and locating on a screen.

In the pretest women did not score as well, which is consistent with past findings. And the end, women had gained more, and had almost equalled the men — to a non-statistically significant level. Five months later they tested them again, and they had maintained the ability.

Takeaway’; Women can excel at spatial attention gameif you give them the time to learn. How can you get new gamers to invest 10 hours in your game while they improve?

#5: The exit screen matters.

Every year we identify one piece of the game that is undertheorized and underdesigned. In the first downlaod it was failure. Last year it was what happens when you die. This year it was the exit screen.

World-renowned architect, dean at grad school of architecture at Columbia. Took what he knows about real-world space and wrote an essay about how to better design the exit in a game space using the theory of architecture.

“Are you sure you want to quit?” — almost like the game has concern for us at the momentof leaving. in Daikatana it was “Are you broken?” The only real risk is to leave the game.

“Game space is a space defined by the complete occupation of our sense and attention.” — quote from the essay. “Game space is the only space that mobile phones and emails don’t reach. There are no messages from another world because when you are in the game you are in another world.” So he is defining game space as what real world space used to be. In RL today we are in multiple spaces all the time.

Exiting a game is not like turning off a tv or closing a book.
The abruptness reveals how deeply immersed we have been. So the moment of exit needs to ne dramatic and noticeable.
The only real risk is to exit.
Like any other architecture, the real key to game space is to design both the entrance AND the exit.

#4: Musical instrument tutoring can make you a real musical hero.

CS and music scholars at National U of Singapore. I was drawn to this study because of the popularity of Rock band and Guitar Hero, really bringing public awareness to the medium. The common objection is why would you spend all this time learning to play a fake instrument in a fake environment.

There is a disconnect, of course — you are not learning to play guitar but instead in a sim of rock star performance. But it is a valid point anyway, and even if not, it’s a public concern: moms calling in to radio, saying they’d buy their kid a real guitar, but not the game.

Insights; they applied lessons from traditional music education. It’s important for a music student to understand WHY a performance gesture is wrong, not just that it IS wrong. This is pretty uncommon in the music games. It will tell you that your technique is bad but just tell you what your pitch is, not how to get better.

Exercises are also really important in music learning — you practice before you go and perform. Many games offer practice mode, but it is the same as normal mode but with no score. Should watch the player play and suggest exercises.

They split the vision of future work into teacher’s lessons, student’s practice, and student motivation.

They went backto look at Loom as well as more recent games… They took these lessons and made tuning exercises for a violin teaching software.

Takeaway: music games can answer the criticism about why not play a real instrument without sucking and turning into just educational software.

#3: voice chat makes measurably makes you like your guildmates more, usually.

USC Annenberg and U Delaware. Did an experiment with two pre-existing guilds that had not used voice chat. had one use text chat only for a mont the other they sent them hardware and Ventrilo and had them use push to talk VOIP. Then they monbitored and did tests before, during, and after.

How did voice chat impact social bonds? Mixed results.

– It intensifies social feelings. Players had richer relationships: deeper, better and worse. More positive and more negative emotions. If you liked someone, you liked them more. And if you hated them, you really hated them.
– They also found that people who used voice as well as text were more likely to only talk among their guildmates, whereas those who used text were more likely to talk to strangers.
– Over time they measured players’ levels of happiness and they found that the folks who were text only sort of declined over time, got a bit more depressed and isolated. Both guilds had members who knew each other in real life. These were guilds together in ayear, reaching end game, smaller, and realizing they were not big enough to do raids content, so there were natural tensions there. The voice chat folks were better able to want to stick together. (shows graph)

Takeaway: voice intensifies social impact, which can be a mixed blessing. How can you help players mitigate the downsides of voice chat?

#2: there are three specific ways you can increase the monetary value of avatars. U of Oulu Finland.

They wanted to know how are avatars valued — what makes the price move? Took a very rigorous analytical approach, borrowing from Nick yee’s motivation studies. Achievement motivations (advancement, mnechanics, competition),  social motivations (socializing, relationsships, teamwork), and immersion motivations (discovery, rp, customziation, and escapism).

So how can you map these three motivations on avatars to determine value? In achievement we can think of stats, wealth, etc. Can measure social value with things like social connections, # of other player they are connected to and able to play with, member of a guild, have a reputation with value, etc. And the immersive aspects can be measured with uniqueness in the game world, the understanding of that avatar’s story, history in the game, changed in appearance over time, the way they are perceived.

If you add up these three kinds of value, they argue you can actually numerically calculate the value of an avatar. Achievement is easy to measure, the social stuff harder. And even the value of a screenshot of old cherished characters was immense to them. So the economy of avatars is not just stats but also proof of what they accomplished.

Takeaway: youc an increase player investment in your game (motivation to stay, not just sell) but building all three of these sorts of value.
And the overall value of game characters is increasing because of the persistence of game worlds, and also the potential interoperability across game worlds.

#1: the top research finding for 2008.

Videogames are the future of live sports.

Usually we look at aspects of the games research landscape… so doing a full survey was fascinating. And when we started, there was almost no work on sports games despite the huge amount of sports games sold every year. But this year, we have a sports study, and it’s a doozy and broke my mind. Indiana University.

The notion that real orld games are an academic or niche pursuit. In addition to real world games, we could toss in augmented reality into this mix. Still sort of an academic pursuit, frivolous from commercial development perspective. But this approach is of a totally different sort.

They asked, how do we mix real sports events with virtual ones. They discussed how sports viewing is changing in general.


– Net access is pushing TV aside as a dominant way for fans to participate.

– Games don’t let players participate except with rosters and the like.

– But live sports generates a ton of live data ready for transmission.

We often find ourselves on the couch with a laptop while we watch, interweaving the two things.

So there is a trend towards the evolution of sports games becoming a sim of televised sports — more a sim of watching than playing. Use of ads, TV graphics, etc, more and more common.

But if you imagine taking the visual aspects of watching the game, and taking all the live feed info, etc, and integrating it into the game.

There’s a whole world of untapped mixed reality games that aren’t about wonky telepresence and augmented reality, but rather based on all of the data we are producing. The researchers made an attempt to make a prototype using both real multiplayer performance, and also the results of a live local game in which real players were playing.

What they were tyring to do maybe didn’t quite work out — when a team scored in the real world, it affected how the computer players would feel about their performance. So they tried to couple one kind of data to the game.

Takeaway: sports viewing is changing, and videogames have a huge role to play. Ask how your next game in general can take the realities of live sports (or something else) and make a connection with it in gameplay.

So how might any game mix the realities of live real events with a virtual version of it. (Audience mentions WildTangent who did a baseball game with these characteristics).

There is a Shadow Top ten on the website with ten more cool findings on the website at

Overall: what connects all of these? A common thread. In game studies we talk about a magic circle, but more and more we talk about the engagement of a game with the rest of our lives. The circle has been broken.

  31 Responses to “GDC2008: Game Studies Download 3.0”

  1. One of the hardest struggles in building a curriculum is the lumbering feeling of the classroom plan. At the level at which the student experiences it, it doesn’t twist or turn or adapt. Consider Raph Koster’snotes from the Game Designers Conference 2008

  2. The idea that it would take a mere ten hours to bridge and negate a long-entrenched gender-based mental difference is utterly mind blowing. Dude.

  3. Agreed! 10 Hours seems so short.

    I find this to be perfect timing, as I actually read Raph’s view on how players can achieve more by playing games that don’t work with what they are used to or perform well at.

    It hit the point home more firmly having it come from two sources.

    Like the study notes though, how can developers keep players interested in the game for 10 hours if their skillsets are not synced up with that kind of game?

  4. 10 hours only sounds short to a gamer. A 2-hour game session to a non-gamer can be a long time if they aren’t immersed in it, so that would mean a non-gamer would have to willingly pick up a game to continue playing it at least 5 times, which is a lot if they keep not feeling immersed in the game.

  5. See? I’m not the only person at the University of Essex doing computer game research!


  6. Points 9 and 10 are very interesting. I’m reminded of a lecture from TED by Dan Gilbert on real v. synthetic happiness and how, not only are they essentially interchangeable, but that people are often happier when they aren’t given complete openness and freedom. Reminding players of the limitations of their playspace is just as likely to generate real, compelling happiness as shooting for total seamlessness—perhaps even more so.

  7. […] 3.0 (Jane McGonigal, Mia Consalvo, Ian Bogost), points #9 and #10 (thanks to Raph Koster for posting a summary on his blog; visit there to read his description): #10: The best content understands exactly how the player […]

  8. 9 and 10 are possibly style dependent. Breaking the fourth wall as part of the style or story (a la George Burns and Gracie Allen) can be a good technique. It can also be overdone. Even Shakespeare wrote scenes where a character appeals to an audience (Puck is a good example).

    Constraining an environment to a task or orienting to the game is precisely why games and virtual worlds are different expressions using similar means. That is what a genre is. That shouldn’t surprise anyone. What would be interesting to study is how open environments become game-like as users contributing content evolve their own consensus about the closures. What kinds of mechanisms facilitate the recognition and negotiation of the commonality? I’ve seen references to worlds with commons that force encounters, and resistance to the same (rather teleport). It appears to be a meaty area for research. One might note that office cubes were designed with that in mind and in real space, people don’t like them, so there is a balance that has to be achieved and one asks if this a commonality of varies by genre and demographic.

  9. The match-content-to-player-skill one, #10, only lasts in the long term if the players aren’t aware of it. If they are aware of it, fooling it becomes part of the game.


  10. Indeed so, but if the content is procedurally generated, couldn’t the game take that into account and evolve as the players’ awareness evolves? If the players succeed in fooling the game, then the game iterates and makes new content that accommodates and builds on that modified play style.

  11. It saddens me in a way that this much analysis goes into what I think of as a largely creative endeavor. But I guess that’s what happens when “big business” comes in and unleashes their bean-counters on you.

  12. […] The future of the Survival Horror genre…s-download-30/ I was surprised to see certain game you love and adore at #9. Even after almost 10 years, it’s a […]

  13. Chabuhi, actually, these are professors, not funded by big business and beancounters. 🙂

  14. if the content is procedurally generated, couldn’t the game take that into account and evolve as the players’ awareness evolves?

    It’s hard to tell the difference from a player who is doing his best but playing poorly to a player who is intentionally playing poorly.

  15. It’s hard to tell the difference from a player who is doing his best but playing poorly to a player who is intentionally playing poorly.

    Fair enough. Although, I would expect that trends would emerge when the metrics sampling pool gets large enough. If players are consistently breaking the game at a certain point, eventually you ought to be able to form a fairly well-founded conclusion that players have found an exploit, and you could generate a content iteration. I’m sure it would take some fine-tuning and developer oversight, naturally.

  16. There have been plenty of games at this point that adjust their difficulty to suit the player and it’s extraordinarily hard to get right. Arcade driving game designers have a particular fetish for it for some reason, to the point that you can often just dawdle around for most of the race then rocket past everyone on the last lap. As soon as you spot the trick it kills the immersion and the game becomes utterly patronising, or worse, appears to punish anything but mediocrity.

    Better to provide a naturally balancing system like flOw, or even the tried-and-tested levels in a traditional MMO… they keep the next challenge just beyond the player’s current abilities but don’t stop them from ploughing ahead if they want a tougher challenge. Use the dynamic stuff to keep things interesting by providing a range of challenges for all abilities, by all means, but don’t fuck with the player by changing things behind their back. They will notice.

  17. david swift: What we were trying to do when we did the experiments for #10 was not simply to adjust the difficulty levels of tracks, but rather to come up with content that fitted the style of particular players. Like, this guy has very varying success on hair-pind bends, he seem to be in the process of learning how to take them; give hime more of them! (Plus, even “rubber-band AI” where the difficulty adjusts to your skills can function very well. See Mario Kart.)

    david mcdonough: Very interesting suggestion. I would love to have access to all that data and try to identify those patterns, but it’s not likely to happen… However, the stuff Ralf Herbrich et al. has done for matching player skill levels for Xbox Live is quite impressive. Of course, he had access to the data.

  18. @ david swift

    Peter Molyneux gave an interesting (if extremely brief) description of how they handle “scaling” of difficulty in Fable 2. I won’t do Molyneux any justice by trying to paraphrase him here, but you may find his explanation interesting. He spoke about it during a preview of the game at (I assume) GDC — the video is in three parts on, though I’m sure you can find it elsewhere as well.

    In any case, Lionhead seems to have arrived at a sound solution for treating players of different experience levels both programatically and by virtue of the design without adversely impacting the play experience.

    Again, I will only manage to butcher what Molyneux said in the video, so I encourage you to give it a view. I’ll post back here if I can find a direct link to the whole presentation.

  19. @chabuhi: Actually we want them to be doing this. We want to know how to make better games, and (maybe) how to make new kinds of games that will be accessible and interesting to an ever-larger audience. (Read: will sell $$$$$). Years ago the academic world and the game industry didn’t even know how to talk to each other, so its nice how much closer they are now and nice to see academics doing research on practical stuff that might help us make better games. Partnerships between academia and other industries have led to lots of good stuff in the past, and finally the games industry is getting some of the joy from that too.

  20. @ Julian

    Generating appropriate content is definitely the way to go. I’m actually very excited about that and it’s great to have some research backing it up.
    But, maybe this is just the gamer in me speaking, unless you can guarantee you’ll judge the difficulty just right 100% of the time, for every single player, better to let them have some direct input as well. Mario Kart’s “rubber banding” drives me throw-gamepad-at-floor mad. 🙂

    @ chabuhi

    Molyneux is always interesting to listen to. I’ll try to hunt it down, cheers!

  21. Regarding #2 and #3, I think there’s a lot of potential tie-in between voice chat, and numeration…statification…er, making stats of social ties. 🙂 Every ten mutual people on your friends list, you get a reputation point. Everybody who has your voice chat ID gives you another point. So many points, you get a title and a colorful beanie. It’s starting to sound like MySpace now, but with more game….

  22. I’m too much of a contrarian to play Slyfiend’s game. I’d make it a point to keep my Social Points at 0. 😉

  23. david mcdonough>If the players succeed in fooling the game, then the game iterates and makes new content that accommodates and builds on that modified play style.

    Then the player will game that iteration process, too, and so on indefinitely.

    If players realise that content is being produced for them, then they have to think: OK, so the challenge of this game has been removed, so is there anything else that’s keeping me playing? Now with a car game, the thrill of the ride could well be enough – after all, rollercoasters are fun and there’s no gameplay involved in those at all. A rollercoaster that modified itself to optimise the rider’s experience would be great! Likewise, a car race with an exciting ride could be just what you want. Faced with a choice of putting your foot down and going off like a rocket up a straight or steering yourself through a corner, which do you do?

    Some people like Oblivion’s level-adjustment system, so when you come across bandits to start with they’re armed with weapons made of wool and then 20 levels later they’re armed with plasma beams (well, Oblivion’s equivalent). This makes levelling up fairly pointless, but if there are other reasons to play (eg. you like the game world) then that can be enough – and indeed it is for many people. As a game, though, it sucks.


  24. I’m too much of a contrarian to play Slyfiend’s game. I’d make it a point to keep my Social Points at 0. 😉

    That could be part of the game too. Keep your social at 0 for long enough, and you get a long black coat and emo haircut!

  25. “how to better design the exit in a game space using the theory of architecture.”

    Man, I’d love to do a game about J.P. Sarte’s No Exit, be pretty fit for this concept.

  26. […] session I wish I hadn’t missed: Game Studies Download 3.0, liveblogged by Raph Koster and revealing “the ten most interesting research findings of the […]

  27. […] short description and some discussion on Raph Koster’s site: the paper: surprising to see that I’m doing […]

  28. How can you get new gamers to invest 10 hours in your game while they improve?

    Answer: not have them be killed or griefed in the first frame. That’s what makes people, especially women, give up.

    Watching how people use space in virtual worlds, I’ve noticed that the men and the TVs always march in and go all around the space, the periphery, flying or walking rapidly, territorializing it, seeing what it is. That’s how I can tell the TVs are real-life men as they do all that rapid walking to and fro.

    Women tend to come in and sit. Do they like the camera angles? Are the ceilings too low? Is there a kitchen? (Yes, people want them, even if they can’t cook real stuff in them; they just “like” them because they have fun decorating and standing in them). So they don’t familiarize themselves with the whole space rapidly such as to have a kind of mind map; they have more a content inventory and an emotional report.

    That’s how I explain these findings about the ability to win these FPS games quickly or not.

    I think these behaviors are somehow hard-wired, as politically incorrect as some may find it. Perhaps you can compensate against them by training as this study suggests.

    I think to understand better the next step, you’d have to study why women give up in those first frames, because they are being killed, ganked, griefed, not understanding the controls fast enough, not seeing what shoots where, etc. etc.

  29. […] of the year. The main question is what do we know about games and game players in 2008 that […] Read More… Published Wednesday, February 20, 2008 5:44 PM by Raph’s Website Filed under: Game […]

  30. […] A sho&#114t &#100esc&#114&#105pt&#105on an&#100 some &#100&#105scuss&#105on on Raph Koste&#114’s s&#105te:http://www.r&#97phk&#111&#115ter.c&#111m/2008/02/20/&#103dc2008-&#103&#97me-&#115tudie&#115-d&#111wn… […]

Sorry, the comment form is closed at this time.