Michaël & Auriea @ the Game Developers Conference Europe 27-29 August, 2002, Earls Court Conference Center, London, UK.

This was probably the first time in our lives that we have attended a conference without being one of the speakers ourselves. We were complete virgins in so many ways. I'm not sure what we were looking for but we have certainly found a lot. Especially as a result of discussions amongst ourselves. Discussions that were sparked by both the lectures we heard and the general atmosphere.

The theme of the conference was "Make better games". And from the subtext of what many speakers were saying, we could deduce that a lot of developers are seeking the ultimate game recipe. They seem to be desperate for the secret formula to making a better game. But if there is one thing that we have learned from this conference, then it is that there is no such thing. Suitably, the conference's closing panel discussion made a joke out of the very idea of "Choosing the Best Game Ever".

Contrary to what one may deduce from discovering the plethora of franchises, sequels and clones that are currently available on the games market, we think the offer is about to explode in many different directions. Sure, there will probably always be shooters, sports simulations and role playing games but more and more games will appear that do not fit in any category. In fact some of these games might not even be considered games anymore, while others might be so in only an extremely infantile way. Even the game developers themselves, with their quest for movie-realism and thick plots, still seem to stick to Tetris and Mario when it comes to naming their favorite games.
It is surprising then to hear that nobody at the conference seemed to be interested in making smaller games that require lower investments, therefore involve less financial risk and more creative freedom thanks to the reduced commerical pressure. They seem to be all still after the Hollywood-scale games that are suffocating both creativity and business in the industry, in our humble opinion. The big game is part of the cherished illusion that people only play one game for a long time. And while this may apply to a small group of "hard core" players, we think most gamers have a life apart from playing games and probably play many different games according to which mood they are in. A developer should therefore probably try to make a game that complements the life of the consumer in some way (rather than replacing it...).

Apart from the start and end panel discussions and two keynote speeches, there were three sessions of five talks on Wednesday and Thursday. Each time, one of those five talks had to be chosen to attend. We decided to always attend them together because we consider the discussions they sparked between us, to be more important than the actual information they were giving us. The three speakers that made the biggest impression on us, Richard Evans, Peter Molyneux and Jonty Barnes all came from Lionhead, the studio that made Black & White, a "god game" in which the player converts disciples and educates a creature. There seems to be a very healthy open atmosphere in Lionhead that allows for some very interesting topics to be investigated and discussed.

Some 500 people must have attended the conference. About ten of them were female, two black, twenty oriental and over 200 wore glasses, half of them oval combined with a goatee. Almost all of them wore t-shirts and jeans (the latter often tan or olive, the first mostly black with a game-related graphic). Just in case you're interested...

Wednesday, August 28th

"Where is Technology Now?" - Panel discussion with James Binns, David Braben, Demis Hassibis, Colin Hughes, Peter Molyneux, Jez San.
Typical, isn't it? When game developers are made to think about "Making better games", they immediately think of technology. The panel came to the conclusion that the future would see more larger developers and less smaller ones as a direct result of technological evolution. Since new games technology will be able to render much more detail, all that detail will need to be made. The making of all these things is still a very laborious task and will involve more and more people. This will make developing games even more expensive than it already is. Peter Molyneux expressed this acutely in stating that developers will spend most of their time "modeling dust." Jez San however, was skeptical believing that with future hardware perhaps a few smart animators can remove the need for an army of animators. The current popularity of "retro-games" did not lighten the mood of the panelists.
Network connectivity was also addressed but the focus was more on distribution than actual gameplay. And someone mentioned voice recognition as a great innovation and he gave an example of yet another combat game with the difference being that you could order your squad around with voice commands. Some of the panelists mused that perhaps less hardware means more fun for players while at the same time they are facing an immediate future of headsets required for players to "talk to" in game characters with the upcoming XBOX online initiative.
In the Q&A that followed that session someone from the audience asked the panel what they thought about Doom 3 (FYI: Doom 3 has not been released yet, but it's innovative 3D engine is making a big impression in the games world). Again it was Peter Molyneux's comment that struck us most: "Doom 3 scared the bejesus out of me. Which at the end of the day is what it's all about."
We walked away with the feeling that despite the fact that developers stress the importance of gameplay and content, they are still addicted to trying to display as many polygons on screen per second as they possibly can. It's like they know they should care about the game but they can't help caring so much about the technology.

Keynote: "The Method: A Model for Game Design"
Mark Cerny talked about a new model for the game development process that stresses a long pre-production phase. It struck us that what he proposed as an innovative new "Method" was actually surprisingly similar to our own schedule for "8". So that was reassuring.
He method boils down to a two-fold process. The first part is the pre-production phase and the second the actual production phase. The idea of the first one is to completely design the game and, at the end of it, to deliver what he called a "publishable first playable", meaning a demo of a few levels of the game that looks and works exactly like the finished product will. This system would require making 2 to 5 prototypes with a core team of the best people in the company over the course of one year. This phase would cost one million Dollars. And at the end of it, the publisher who coughed up the money, would still have the choice of actually publishing the game or keeping the intellectual property (or selling it to the developers). The main advantage for the publisher is that the high-risk phase of the project would be reduced to the pre-production phase.
Mr. Cerny completely got rid of the notion that it was required to write the typical 400 page design documents that developers tend to make, a document that contain each and every detail of the game, in favor of a five page concise game description which gives the publishers an overview of the game. (This was later confirmed in the panel with the publishers.)

Lunch - "Avatars Offline : The MMORPG Documentary"
This was a documentary movie about "Massively Multiplayer Online Roll Playing Games" like Ultima Online and Everquest. It was frightening to see how addicted some people are to these games and how much time they spent in the virtual worlds they create. At the same time it was fascinating to see the impact that virtual and mixed realities had on the course of their lives and how "just a game" becomes something quite real. Something for game developers to remember when inventing these environments is consideration for how their game may be morphed into something quite unexpected once the plots are brought to life by the minds of gamers.

"Scripting in BLACK & WHITE: We are not Making Movies!" - Jonty Barnes
Black & White is one of our favorite games. And we like to differentiate games from movies. So this title attracted us.
In Black & White, gameplay is sometimes interrupted by cut scenes. These scenes are not pre-rendered movies, but play in real time in the game world according to pre-written scripts. As a result the weather in the movies may be different, or some characters may appear on screen unexpectedly, etc. Jonty Barnes talked about the pros and cons of the design and technical decisions they made to integrate these movies with the game. The camera system they had developed, e.g., turned out to be too simplistic for use by real cinematographers. He also advocated allowing some freedom while recording the voice sequences (also in translation) and doing those before animating the characters.
We were a little disappointed to hear that "Black & White 2" was going to be all about war. Mr. Barnes explained that they wanted to add more drama to the Black & White concept and illustrated this by playing yet another Carmina Burana sound-alike while showing a Black & White landscape. As if there wasn't enough war and "epic tales" in games already. The first Black & White already contained a war-story in the shape of the battle between two gods. This was the weakest part of the game, in our opinion. But, just like Black & White, this game will probably contain many different games within it, some of which we like more than others.

"Getting Data that Improve Games: A Case Study of HALO" - Fulton/Steury
We haven't played Halo. It's apparently a first person shooter on the X-Box. The images of it that they showed during the talk could not convince us that this game was any different from Half Life, Counterstrike, Deus Ex, Oni or any Quake-clone. But the talk was not about "Halo" as a game, but about the process of user testing that went on during it's development. This was very blantant advertising for Microsoft's services.
The decisions that those people took based on "experimental psychology" performed on 6000 Seattle gamers could have been taken by any designer who knew his job. And still the resulting adapted game would have baffled us as a player. From the looks of it, "Halo" struck us as one of those games of which we wouldn't even be able to get through the tutorial of... too many band-aid fixes based on design by committee.
In spite of this the speakers did give some interesting models to follow in terms of how game testing should be conducted early in the game design phase. They put an emphasis on doing testing in an almost scientific manner and atmosphere with testers making efforts not to influence the reaction of the test subjects.

"Making Sense of Gameplay and Emergence" - Jesper Juul
Mr. Juul is an academic. Games are a hot topic for hip academics these days. Most of them completely destroy the delicate embryo of a medium with big elephant's feet that set apart games from non-games by judging whether what you did in them was actually playing. And they usually define playing in terms of chess and board games as part of the trend of saying that you focus on gameplay rather than fancy graphics or narrative. While maybe the problem with games is that they are simply ugly and do not have very interestiung stories to tell. The academics do not seem to understand that "game" might simply be a badly chosen word for this form of entertainment.
Lucky for us, Mr. Juul was not so aggressive. In fact he pointed out that the popular notion of a game being "a series of interesting choices" does not hold up. Analyzing Donkey Kong, he pointed out that a game consists of many un-interesting choices and in fact only a few interesting ones. Yet both are fun and are part of gameplay.
He also offered a definition of emergent properties, games made as a system in which players can adapt their gaming strategies to suit their own style of play. He drew a distinction between emergent play and simple exploits in combinations of games rules. He pointed out that most non-computer games have emergent gameplay and blamed linear storytelling for the loss of emergence in computer games.

Thursday, August 29th

"From Pitch to Publish: Getting the Deal" - Panel with Kevin Bachus, Zeno Colaco, Adrian Curry, Jason Della Rocca, Francois Masicopinto
To our surprise, this turned out to be one of the more uplifting sessions of the conference. We had expected to learn how evil publishers were and how they completely destroy all innovation. But in fact they turned out to sound like they were pretty open minded about things and allergic to burocracy. They disapproved of developers presenting their game in huge design documents and defining them based on other games ("Our game is like Grand Theft Auto 3 meets Tomb Raider"). Grand Theft Auto 3, by the way, being the game that developers throughout the show seemed to appreciate a lot. They also made a plea to those coming to them to pitch their game concept to hold back on the descriptions of game setting in favor of talking about more practial aspects of the game such as how will this game appeal to consumers. On this subject Mr. Masciopinto, of Infogrames, lamented that if one more developer comes in and starts their pitch by saying "It's 5000 years in the future, you are on an alien planet..." he was gonna scream.
Jason Della Rocca from the International Game Developers Association had done a little email survey among game developers and publishers and presented us with the results in these nice little lists:

What publishers want:

a "hook" = unique selling point

track record/company in place

clarity & focus

interactive demo

What developers forget:

a "hook"

competitive analysis

this is business

clarity & focus

to be honest

What publishers don't want

huge design documents

platform specific technology features

non-interactive demos

original Intellectual Property

The other panelists, two publishers and two console company representatives, seemed to agree with most of this, accept for the latter point. The publishers insisted that they did want original ideas. Kevin Bachus argued that the games industry is losing touch with its audience and that publishers are trying to reduce the risks by investing in sequels or derivative products rather than original games. The only way to get out of this dead end is to start listening to consumers again.
In general the publishers expressed a desire for a partnership with a developer based on a shared passion for the game. This may seem obvious and might sound suspicious coming from a publisher. But I wouldn't put it past the developers that we saw during the conference, of being complete weasels about the kind of things they make. They also all seemed to agree that it really helped to form a personal bond with someone at a publishing company, to find what Adrian Curry referred to as a "champion". Someone who could be the person established within the industry to endorse you, help spread your name around and influence the company to take a chance on your game idea.

"Systemic Level Design for Emergent Gameplay" - Harvey Smith
Mr. Smith talked about the transition from building levels for Deus Ex on a per case basis to setting up a library of reusable items. In fact he explained to us the advantages and disadvantages of the age old concept of object oriented programming, as somebody in the audience remarked.
He seemed to be very excited about the resulting emergent gameplay that allows players some freedom by making unexpected combinations of game objects like blowing up robots to open doors and the like. He warned for less desirable forms of emergent gameplay like the one in Deus Ex where a player could use mines to build a ladder on a wall that he could effectively climb and from which he could see parts of the game world that were not intended to be seen.

"Social Processes: Moving Beyond Individualism" - Richard Evans
A lot more interesting to us was Lionhead's Richard Evans' talk in which he presented a new model for independent behavior of Non Playing Characters (or NPCs: the characters that are part of the game world and that do not represent the player) in games. This model is based on the presence of invisible "social process clouds" which request the character to do a certain thing. The character in turn responds after evaluting the repercussions of the act based on the other social process clouds that surround him (often hundreds at a time). This model would be much more efficient than having to set up a whole library of actions for each individual character.
This idea was based on concepts developed earlier by philophers like Wittgenstein and Heidegger, people who, as Mr. Evans pointed out, were very suspicious of the very concept of Artificial Intelligence. But from where Mr. Evans was standing, there was a one-on-one relationship between their theories and the practice of game design. Or as he put it: "Take the hippy philosophy and type it in!"
The reason why this model is so interesting is that it expands the socalled "verb-set" of a game character dramatically. The "verb-set" is the amount of actions a character can do: run, jump, climb, etcetera. By connecting behavior to social environments, characters can appear much smarter and the world much more alive than it or they would otherwise.
In a way this connects to Harvey Smiths application of object oriented programming to game design from the previous talk since the result of this "systemic" approach is "emergent behaviour", a buzz word in contemporary games development circles. It means that unexpected things can happen when you define objects by a set of rules rather than on an individual basis. Mr. Evans gave us an example of two groups of NPCs that started fighting each other without the game designer having intended that, simply as a result of the social processes and the characters' responses to them.
In an answer to a question of ours after the talk, Mr. Evans admitted that all of these processes were being set up by hand. The next step, in our opinion, would be to break down these processes to a small group of constituent parts and let the processes build themselves by combining these parts. Anyway, we can expect lots more exciting stuff coming from Lionhead...

"Beyond A Great Game, Building a Franchise" - Bruce McMillan
We were in doubt. Do we think Mr. McMillan is a complete and utter asshole or do we admire him intensely? He is a classic case successful yuppie, has been working for what must be the biggest games publisher in the world (EA) for all his life on derivative games (FIFA soccer and Harry Potter, to name a few). But the man is so passionate about making money that it is almost moving. He advocated making games based on already existing stuff like books, films or other games (sequels) because then the consumer would recognize things, code would be reused and the creator would get more freedom (funding) and the ability to perfect their previous work. It's a good thing that we see "8" indeed as part of a series. One day, Mr. McMillan will be very happy with us. But we shouldn't be so harsh. He truly seemed to be concerned about the quality of his work. His only mistake is that he lets the audience be the judge of that in terms of sales figures.

"Choosing the Best Game Ever" - Panel with Owain Bennallack, Charles Cecil, Martin de Ronde, Jon Hare, Phil Harrison, Mike Goldsmith, David McCarthy
This turned out to be a hilarious election of "the Best Game Ever" by the audience booing and cheering after the panelist had each defended what they said they thought was the best game ever. With the exception of Grand Theft Auto 3, those games were mostly old games (Elite, Mario, Tetris, Space Invaders). When the audience was voting, we couldn't help but realise that none of the games they suggested could get our vote. In fact we couldn't think of any one game that would deserve that title because that game has not been made yet. Some games that exist now carry a lot of promise but none of them is "the best ever". And we certainly don't like old games from the arcade period as much as most game developers seem to. Maybe we don't like games at all...
This realisation did make us feel a little guilty. All these developers have been playing games since they were in their diapers and they all seem to be into it so much. And while games have always been a major influence on our work, we have never been so fanatic about them.
Anyway, we were very glad that, after a whole conference of searching for the magic formula, choosing the best game ever was interpreted by the panel as a farce. This confirmed the conclusion that we had already come to ourselves by that time. That many different games are possible. That they all have their value and that some players play one game now and another game at another time when they are in another mood. There is not one greatest game and there is no need for one. There is a place under the sun for many different types of entertainment.

Things we have learned from the conference:

1. We may not be making a game but that's ok.
   Maybe what we are making is a kind of poetry. We want people to "read" it and then walk away from it with the memory. We want our world to be available to them at any time, but we don't want them to be addicted to the game.
   The gameplay in our game is really just an excuse to get the player from one narrative element to the next. It is ironic that many games are exactly the opposite: in most games the story is the glue between the sessions of playing.
   While we do like the playing in games, it is not our ultimate basis for judging them. I guess we look at games as works of art. We do enjoy the brush strokes and the colours but for the painting to make a lasting impression, something else is needed, something that may be very hard to define and may better be left to intuition and instinct.
   
   
2. We may need to approach a developer rather than a publisher with our ideas and demo.
   We are a small team. And even if we incorporate, we will be a small company. Publishers do not have the resources to put together a team for us. And we may not have the know-how to hire people to work with us. Partnering with an established development studio may be the only way of actually producing and even publishing our game.
   This of course depends greatly on the outcome of the research project we are involved in right now. While we may currently be in a phase in which the project seems too gigantesque for us to deal with on our own, we will not take any decision until May 2003, when we will have a more realistic perspective on things.
   Part of the project has now become figuring out our preference: will we limit the technological scope of the project in favour of more integer content but taking the risk of never seeing the game distributed properly or will we search the help of a third party that can help us put the product on a "professional" level, risking losing control over the integrity of the game in favour of better chances with publishing. The first option seems to be very risky but the second requires a serious change in life style.
   
   
3. We need to talk about our ideas with other people from the industry.
   It is perfectly possible to change many elements in our game design without touching it's core. It would be very interesting to discuss our ideas with people who have more experience than we do.
   
   
4. We should try to write a description of our game that captures it's strong points, or at least a list of our objectives and priorities.
   This is important for us in order to present the game to both players and publishers/investors. But is also a way to keep focused. It is very easy to get off track while working on a project as complex as this. We could easily spend the rest of our lives on it. But while that prospect may have it's charms, it is not a realistic option.
   We have to decide where our priorities lie and then work based on that: spending more time on high priority issues while reducing the time wasted on things that ultimately don't really matter. And we should probably not make a distinction between what is important for the demo that we are making and what will be important in the final game. The demo should be a good game in and of itself, albeit a very small/short one. We should think of the publishers/investors as players.
   
   
5. If our main goal is to move the player emotionally, then we need to work on the character of the girl and her presentation to the player.
   This will require study and discussion.
   
   
6. There is nothing wrong with narrative as such in a game. But we are trying to have narrative that is a. more interesting and moving and b. shared with the audience in a less linear way. While we are incredibly interested in generative systems in games like the one Richard Evans proposed, this is not what we are doing in this game. At least not in terms of the real assets in the game. The "emergent gameplay" that we are going for is in the player's head. While our game will technically be more or less "canned", the possible interpretations and associations should be endless.

conference site:

http://www.gdc-europe.com

related:

Games push for more than pretty faces

Conversations From GDC Europe: Mark Cerny, Jonty Barnes, Jason Kingsley

Conversations From GDC Europe: Bill Fulton, Zeno Colaco, Harvey Smith

Slides and Papers