Molyneux: Industry must find new ways to work

Fable creator 'frightened' at how late in the development process games are actually playable

One of the industry's most enduring game designers, Microsoft Game Studio's Peter Molyneux, has told that the current standard process for creating triple-A console games "has to change".

Speaking in an interview at this year's Develop conference in Brighton, UK, Molyneux - whose Lionhead team is currently hard at work finishing off Fable III in time for release later this year - explained that developers need to work in different ways if good quality games can continue to be delivered to consumers.

Referring to Fable III, he revealed that work on the project was currently "very, very tough," primarily because of the scope of the game.

"Because it is part-simulation, part-roleplaying game, part-action-adventure... it's got more voice acting, a bigger cast, more musical scores, a bigger, freer world - more so than any other game in the whole of Microsoft Game Studios' portfolio, full stop," he said.

"It's got more bugs, more active bugs, than any game that MGS has ever had in its history - and the team of over 100 people are working insanely long hours, and they're coming down. That's the way it is."

But that situation is common with many of the top game experiences out there - an unhealthy situation that isn't sustainable.

"Here's the issue which I hate, Phil - again, we're in the same position as many, many other developers, if you read the post mortems of Uncharted 2 or any game, they all say the same. We find our game so late on in the process, that it's very hard to pull it all together - and that has to change.

"We cannot do this - we can't keep turning up this late. A film analogy is me turning up with a camera on set and saying: 'Okay, I'm not sure what the story is, but let's turn the camera on anyway.'

"We've got to stop doing this, because one, it's too expensive, and two, our consumers, the people that play our games, are too demanding of the quality we have to deliver. We just have to work on a different way - it's got to be the case."

Quite how the industry finds new ways to work is unclear, given the sheer amount of data and technology that goes into each big retail product - but while the industry has mostly moved away from the practice of releasing a game before it's finished, simply waiting ever-longer or drafting in ever-more people also isn't the answer, says Molyneux.

"Well, it's not. It's very difficult - the huge supertanker that is the launch platform is very hard to switch off once you get past a certain point. You've got to book your TV adverts, your creatives have to get together and do your advertising campaign - that happens way, way before you know whether you're really going to get to the right quality bar.

"And again - I just feel that the way that we developed Fable I and Fable II - and a lot of our games - all of our tech reaches up to a certain line when we can actually play the game... and that line seems to be so late on, it's just frightening."

The full interview with Peter Molyneux will be published on in due course.

More stories

Ninja Theory bosses explain decision to join Microsoft Studios

"We want to take bigger creative risks, and create genre-defining games without constant threat of annihilation," says creative lead Tameem Antoniades

By Haydn Taylor

Former Tomb Raider boss joins Xbox leadership team

Darrell Gallagher now at Microsoft, teases big things for E3

By James Batchelor

Latest comments (16)

Ian Wilson CEO, Emotion AI11 years ago
The answer, of course, is and always has been middleware. Stop building tech and just build games.
0Sign inorRegisterto rate and reply
Phil Elliott Project Lead, Collective; Head of Community (London), Square Enix11 years ago
I'm not a developer, but I suspect one response to this might be that different middleware tools are designed for (or better at) different things. Clearly some engines have proven themselves more diverse, but at least one platform I can think of that was basically designed to allow developers to do pretty much anything hasn't made the impact that it might have principally because of that point...
0Sign inorRegisterto rate and reply
John Donnelly Quality Assurance 11 years ago
Another solution could be to look at how games are tested compared to what is going on in mainstream IT. There are a number of automation tools and its common for testers to have titles like QA Engineer.

In any other industry your QA team is seen as a high skilled professional group where certification and training is the norm.

It could form part of a overall revamp of the ideas to solving these problems but the QA aspect should not be overlooked as it is a key service to getting the quality bar right in any game coming to market.
0Sign inorRegisterto rate and reply
Show all comments (16)
Both points are pretty valid. Middleware offers some amelioration of some issues but it still incurs development costs and has its own host of gremlins/issues requiring more middleware support.

Ultimately, if you want to develop a next next-gen game that is fit for purpose then developing your own R&D towards a engine robust yet diverse enough for say:

- unique in-game cinematics
- mass crowd scenes
- Unique AI toolset
- proprietary destructible environments
- proprietary shading, texturing and polygonal limitations
- next gen MMO vs FPS vs Simulation platforms

then ultimately, your own engine might suit you best for years to come (and you never have to pay for licensing) and if you do well enough, you can make it semi commercial or roll it out across your sister studios.
0Sign inorRegisterto rate and reply
Ivan Pedersen Lead Art, Geomerics11 years ago
Make smaller games with higher quality? I'm a fan of the Fable games as they do story and narrative very very well; it's a very good mix of adventure and fun. I am (and never will be) a big fan of games that tries to be too many different things at once. I was somewhat disappointed by Butal Legend's descent into RTS towards the end; got a bit lost and bored playing Just Cause 2. The open-world-free-roaming-I-can-go-everyhwhere experience has never really spoken to me. I'd rather have a small, well crafted and carefully throught out rollercoaster than a large expansive open world filled with mediocre assets, story and gameplay. (not referring the games previously mentioned)

Edited 2 times. Last edit by Ivan Pedersen on 16th July 2010 11:51am

0Sign inorRegisterto rate and reply
Josef Brett Animator 11 years ago
Are Lionhead using the same engine as they did for fable 2? To make a new engine and a new game in 2 years is a big ask.
0Sign inorRegisterto rate and reply
James Watt VFX Artist, Codemasters Guildford11 years ago
We need to create a new model for project development. Many projects go into development with only a rough outline.
If we would follow a Pixar model, where you take some of the key players (gameplay, art lead, programming, story, etc.) and put them in a room for a couple of years to really flesh out a project, a project can go into production with a concrete start.
Companies feel like they should have their best people working in production on actual projects, whereas it makes more sense to have some of these people designing the next title.
We need to have most of a game designed before everyone starts working on it. When a game goes into production, the focus should be on polish and finessing the core idea.
If we looked ahead of just the current title and planned for the future, we will create better games.
0Sign inorRegisterto rate and reply
Well, the difference also probably lies in the power of the director/producer for the film industry. These guys have a pretty clear vision of what they want to get when they decide to work on it. There is also a very long development process with the script etc. When they start shooting, everything is planned out (there are of course surprises, and more than often studio execs come in), but they do differentiate creation and production and have a true master managing the plan, whereas from what Molyneux says, games are much more chaotic in their development.
Middleware is important (special effects in movies are middleware technically), but there is also a need for true talented visionary "directors" teamed with a technically skilled "producer" who can worry about feasibility. On paper they exist, but I'm not convinced much studios really work like this.
0Sign inorRegisterto rate and reply
Even with the Pixar model, they ran into issues. Eg. some productions were worked on for 2 years and canned to be totally razed, rebuilt, crunched and come out the other side.

The matter of fact is, its a creative process. And sometimes to get the really good stuff, its a matter of iteration and polish.

To ameliorate this, mini incubation aspects could be applied.

eg. Mini torture test/stress tests of game engine, effects, character design/feel and interface, environment and level design (could be based on how martial art films are scripted, on set and figuring how a space can be creatively used and reused) and all these separate elements are worked in parallel with early previs of a alpha vertical slice demo could be produced on the fly from day one.

Thus, even with radical rewriting or change of content, the latest iteration will be all the stronger for it due to polish, polish and multiple layers of gloss, wax and abuse to come out all the stronger on the other side.
0Sign inorRegisterto rate and reply
Michael Emeny Senior Lead Tester, Rockstar Lincoln11 years ago
Here's another angle to consider this from. Perhaps games companies should think about holding back their release dates until they are in a confident position to meet it. Blizzard doesn't seem to have any problems telling us 'it'll be ready when it's ready' and I'm sure they put as much effort into promotion as anyone else.
The trick will be to get out of this idea that Christmas and Easter is the be all and end all for videogame releases. Plan to get the game wrapped up with a rough estimate, work toward that and tweak your plan on the fly. Six months from conclusion ask yourself, 'can we do this'? If the answer is yes then crack and and get it done. Make the announcements, roll out the ad campaign and bring on the crunch (which is inevitable no matter how much planning goes in)
If it looks like you need a few extra months, weeks whatever (budget allowing ofc) then make use of that time and hold back any release announcements. If the game is good enough it won't flop just because it falls out of the run up to Christmas. Plenty of decent titles have sold very well outside of Quarter 4. I'm sure this is all very naieve...
That said, I think a company does need to manage its crunch with care. There have been a number of talking points around this topic over the last year or so. This point doesn't really add much to the discussion but it's encouraging to know these issues are in discussion.

Edited 1 times. Last edit by Michael Emeny on 17th July 2010 7:17pm

0Sign inorRegisterto rate and reply
There is one slight challenge with this method.

I suspect a large majority are not independently well funded and require cashflow relative to milestones. And thus delays of any significant amount means more work with less or no cashflow = disaster/closure for some
0Sign inorRegisterto rate and reply
Heinz Schuller Art Director / Artist 11 years ago
James and Shann nailed it. While developing a game is partly about exploration, nothing could have a bigger impact on schedule management and crunch than a true, thorough pre-production cycle. An incredibly rare beast in this industry.

The rub? Many studios look to transition their full production teams onto a new project once the existing project is finished. But the timing rarely works out, using your full staff to finish a game means you haven't pre-pro'd your next project to the degree that it's ready, if at all.

IME the waste and inefficiencies are usually incurred at the front of the project, with an overly large team prototyping ideas that aren't fully fleshed out yet.
0Sign inorRegisterto rate and reply
Philipp Nassau Student - Business Administration (M. Sc.) 11 years ago
Strcitly economically speaking, middleware is the way to go, even though it might not meet the extremely specific demands at some point. It's about core competences. Knowing how to design a game that's fun is a very specific skill and it doesn't come with technical skills inherently. Creating much smaller teams to create the foundation of a game and iterating design decisions long before going into mass producing models, textures and the like saves a lot of money, just like Heinz mentioned. Designing any technical product comes with creating mock ups, doesn't it?

Ultimately, having to reduce the number of polygons in a scene by 5% shouldn't have an impact as great as the difficulties that arise from creating one's own tech. That, however, doesn't mean that technical details shouldn't be included in the thought process.
0Sign inorRegisterto rate and reply
@ Hienz

How about, for a small to medium sized studio, when production reaches 2/3rd of the completion - set the pre-viz/concept art team paired with 1 programmer to form a incubation pre pre-production team to explore a new IP/existing IP such that when production is complete at least 3-5 months pre pre-production would have taken place. Upon production hitting gold, the team can ease into early pre-production and vertical slice demo, prior to going into full production

For larger studios, 1-2 incubation teams could start a new IP development midway through the production cycle upon completion of most production/conceptual assets. This way, by the time your product A is gold - the new project is already ready for greenlight production and so forth.
0Sign inorRegisterto rate and reply
Heinz Schuller Art Director / Artist 11 years ago
Dr. Wong, your suggestions are sound. But the real trick is keeping that prepro team from getting piled back onto the main project during the last 6 months before ship, when resources are in high demand. :)
0Sign inorRegisterto rate and reply
Andrew Lee Pearson Studying Game Designer, Train2Game11 years ago
Don't know if the consumer would be happy with this, but in order to meet the deadlines not go over budget and all the other business related stuff behind game design. This depends on the genre of game, would people be happy with 2-3 perfect running levels and then a patch unlocking the rest every couple of weeks or how ever long, I know you would get the usual uproar of it ain't finished, but working with consumers stating we had to release it this way rather than give you a fully accessible buggy game we have given you quality levels to get you started with.

Kind of like episodes in a series...
0Sign inorRegisterto rate and reply

Sign in to contribute

Need an account? Register now.