September 23, 2004 4:00 AM PDT

Game publishers sweat console change

As gamers anxiously await powerful new consoles from Microsoft and Sony, game publishers are trying to figure out where they'll get the extra millions of dollars needed to create titles for the new hardware.

With schedules and budgets of top-tier games already approaching those of Hollywood films, the advent of new game machines from Microsoft and Sony is expected to put added pressure on game developers and may lead to further consolidation in the shrinking ranks of game publishers.

"It's a big, big problem the industry faces," said David Cole, president of research company DFC Intelligence. "As you increase the graphics capabilities on the costs so much more to create a game that lives up to the graphics potential. Developers are really going to have to look at the bottom line."


What's new:
New game consoles promise startling graphics and more realistic play. But additional hardware complexity is expected to nearly triple the cost of developing new games.

Bottom line:
While another wave of consolidation among software publishers may result, game makers are investigating new programming techniques to help ease the pain and save money.

Market leader Sony, along with Microsoft and Nintendo, are expected to introduce new console designs within the next year or two. While Nintendo has downplayed the importance of graphics performance in its plans, Microsoft and Sony are promising dramatic increases in horsepower from the next versions of the Xbox and the PlayStation. Xbox 2 will sport an internal configuration dramatically different from the current console, while Sony is promising revolutionary advances from the "Cell" processor that will power the PlayStation 3.

Harnessing that silicon will mean a lot more work for game developers, said David Doak, director of Free Radical Design, an independent developer best known for its "TimeSplitter" games. Characters in games for the previous generation of consoles were represented by a few hundred polygons, the basic units of geometry rendered by game engines, Doak said. Current consoles use a few thousand polygons, and density is likely to jump to hundreds of thousands with the next generation.

"The problem is that the mechanisms for making those characters and textures haven't really scaled to make it easier to create all that complexity, so you just have to do a lot more work," he said.

Current development budgets for an A-list title average around $5 million, Doak said. For the upcoming consoles, "I expect the minimum will be two to three times the current costs," he said.

"We're seeing some publishers teetering on the edge right now."
--IDC analyst Schelley Olhava

That's a nut that some developers and publishers aren't going to be able to crack. The current generation of consoles has already seen the dissolution of several major publishers, including 3DO and Acclaim, and more may be on the way

"We're seeing some publishers teetering on the edge right now," said Schelley Olhava, an analyst for researcher IDC. "It's going to get harder and harder for the small publishers because the development costs are getting so high."

Jack Sorensen, executive vice president of worldwide studios for publisher THQ, said he expects a few game publishers to go by the wayside, but he dismisses speculation the game industry will become a company town run by leading publisher Electronic Arts.

"With every hardware cycle, there's some shakeout," he said. "When I started in the 1990s, there were all these big Japanese publishers who aren't around anymore...But there's no part of the entertainment industry that has fewer than half-a-dozen major players, and I don't think games will be any different."

Staying in the game
Publishers who weather the transition will have to make a number of adjustments, including being more selective about the titles they produce. Doak expects to see publishers narrow their ambitions to the closest the game industry has to a sure thing--sequels and games based on licensed content such as movies and comic books.

"It's going to become much more difficult to develop an original title," he said. "When the cost of development was lower, it was OK to have these occasional moon-shot titles where you really pushed ahead with an original idea. Publishers aren't going to be willing to take those kinds of risks now."

Olhava said the "throw it against the wall and see if it sticks" attitude of game publishers won't fly as development costs soar. "It's a very hits-driven business," she said, with little customer research to understand what makes a game a hit. "The publishers need to know who's playing games, and why, a lot better than they do now," Olhava said.

Even selective publishers will have to make tough decisions about how they spend limited development dollars. Elements such as cut scenes--the animated and often pointless sequences that bridge levels of a game--may need to be discarded as needless luxuries.

"Developers are going to have to be really smart about what they put into a game and what they leave out," Cole said. "They need to have a good feel for what kind of features customers really appreciate. Feature sets have to be justified as a way to increase sales, as opposed to adding to the artistic nature of the game."

THQ's Sorensen agreed, saying graphical glitz is already being overemphasized in some cases. "The main thing is to focus on entertainment value," he said. "I remember 10 years ago, if a game wasn't offering 50 hours of entertainment value, you wouldn't even consider it. Now it's down to 10 hours for a lot of games, because so many resources are going to the graphics."

Time to recycle
Developers will also have to get better at reusing code, a practice traditionally shunned by game artistes. "We used to be an industry where people wrote the technology on an ad-hoc basis with each new game and then started over with the next game," Doak said. "It's getting less and less practical to work that way."

For some developers, that will mean an increasing reliance on middleware, development tools that apply prescripted code to govern game elements such as physics and artificial intelligence (AI). Electronic Arts signaled its intentions earlier this year by purchasing Criterion Games, whose RenderWare is the most widely used piece of middleware in the industry.

"If I've made the perfect beer glass, why should anyone want to make that again?"
--Fred Skoler, COO of Whatif Productions
EA representatives declined to comment for this story, but company executives have characterized the move as part of the publisher's strategy to deal with the complexity and cost of developing for new consoles. "RenderWare is significant for EA in terms of getting ready for the next generation of consoles," EA Vice President Jeff Brown recently told BBC News.

Microsoft is promising similar advantages from XNA, the flexible game development architecture it announced last March. XNA will give developers unified sets of tools for working on games for current and future versions of Xbox as well as Windows PCs, said Cameron Ferroni, general manager of Microsoft's Xbox Live online game service. Microsoft's objective with XNA is to trim enough trench work from game development that creating games for the next Xbox won't be much more expensive than for the current model.

"Our goal is that as systems get more complex, more powerful...we can reduce the amount of money you have to spend on physics, AI and things like that," Ferroni said. "We're providing developers with additional tools to really speed up the process, and we're making it possible for other people to create and share tools and engines that fit on top of the XNA environment.

"If someone writes the best physics engine in the world, they can put that out and it's there to incorporate in your game," Ferroni said. "We want to make it where games really distinguish themselves in the creative aspects--the artwork, the audio, the gameplay--not the plumbing."

Start-up Whatif Productions hopes to take the reuse concept one step further with a new game engine that allows developers to reuse code between games and platforms. Create a nifty ball-handling effect for an Xbox basketball game, and you can port it right over to the PlayStation version or to another title altogether.

Fred Skoler, the company's chief operating officer, said he is confident the complexity of developing for next-generation consoles will promote interest in reusing code. "We think that down the line, the idea of being able to quickly assemble content and then tuning it to make it your own is going to be seen as having a lot of value," Skoler said. "You've solved a huge problem with this industry if you can start from a very capable foundation and then focus on the creative things."

Skoler imagines a fluid marketplace where developers will be able to grab ready-made components instead of creating their own imitation. "If I've made the perfect beer glass, why should anyone want to make that again?" he asked.

Skoler believes publishers will also appreciate the opportunity to get something out of a game that flops in the market. "The way things work now, if a game doesn't do well, you've basically spent a lot of time and money creating content that doesn't have any value at all," he said. "What if everything you put into the game had a reuse capability?"


Join the conversation!
Add your comment
Terrible idea
I think that alot of stuff discussed in this article is a terible idea. I want every new game I play to be unique, not a recycled pile of code. I can't honestly believe that a physics and AI engine genericaly developed would work as well on a FPS as it would a car game. And I can say that I wouldn't want a Gran Tourismo style game to handle the same as Mario Kart.

I think these companies are looking for the cheap way out. For me, a game needs to be fun. Period. I don't care about fancy graphics, because after playing for 10 minutes you forget about all that, and it is the substance of the game that realy matters.

It will be a shame to lose smaller development companies, as innovation will be exchanged for polygon count.

My 2 cents . . .
Posted by (2 comments )
Reply Link Flag
It's not that bad
The physics engine used in Gran Tarismo could easily be tweaked with a very few changes to make the game totally Mario Cart.

I've played with the physics settings of many racing games and just some small tweaks will make a world of difference. There were some real sweet tweaks that coule be made to Midtown Madness to make a car have super-glue traction for example.

They certainly won't use the AI engine from NBA Shoot out 2010 for the Grand Theft Auto title that follows it to market.

The bad news is they do already re-use a lot of code, it just isn't as structured as it needs to be so that the work can be salvaged quickly and put to use again in a new product.
Posted by (46 comments )
Link Flag
Unique Games, Reused Code
Just because the code was used before doesn't mean that it's not unique. There have already been many bad examples of reused engines, but it's the game itself, not the code, that matters. Modded games can be completely different from the original. If every game had the same processing engine, but different textures, guns, etc, you might not even notice that they share the same engine.
Posted by (1 comment )
Link Flag
How did it come to this?
During the 16-bit generation, when half of the current crop of developers were just starting out, usually as 2 or 3 guys working from home, they used to be able to get a decent bottom line from the sale of 20-30 thousand copies of their games.

These days, for the same purchase price, were buying games that 20 or 30 developers have had a hand in.

So unless Microsoft, Sony and Nintendo want the console business to stagnate, innovation to die, as people finally get tired of the constant stream of RTS, FPS, TPS games continually flooding the market - they will have to create systems (and development kits) that every day joes can code for.

If it wasn't for the fact that systems like the Amiga or the Atari ST enabled sometimes just one person to create, in a few months, a blockbuster game, then companies like Electronic Arts would not exist.

The next generation Sid Meier or Peter Molyneux will never be seen unless the creators of these boxes make it possible.

So instead of just designing a console with the most powerful graphics system, dolby 9.1 and gigabit broadband capability - they need to worry about who can realistically (i.e. in 2 years or less) produce a game for it.

The only way for this to be possible is if development kits can be bought to run on PCs (or Macs for that matter) for under $100. Because by nature bedroom programmers do not have $1000s to spend on systems and software.

Remember that nearly every genre you love today was invented by a company that started out as a group of penniless programmers on sub $400 computer systems.
Posted by ajbright (447 comments )
Reply Link Flag
Microsoft had the best idea... GaSp!!
Really, microsoft actually had the best idea for game developing for a next gen system.. Which was the xbox incubator program.. You'd litteraly lease an xbox development system and special pc hardware and software for a 6 month time period to come up with a prototype.. If the A$$es at redmond thought it had potential, then they let you keep the stuff for another 6 months for free.. then reevaluate and if you had something good going then you'd be a game developement studio funded by M$.. That's the only smart thing they've done in response to xbox development.. Remember that iD software was once 2 guys under the thumb of appogee software doing the same thing with shareware.. :-)
Posted by nzamparello (60 comments )
Link Flag
Code Reuse
I'm shocked that code reuse isn't in wide practice in the gaming
industry already. Why would any developer want to write
inflexible code? I find it hard to believe that developers don't
have a toolset that gets pulled from and added to with every new
game. Is this really true?
Posted by (15 comments )
Reply Link Flag
Code Re-use
Actually code reuse is VERY common... Doom, quake, and the unreal engines have all been used very much so in the video game world... Now the GTA3 engine is currently getting passed around... The first code tool passed around was actualy by Core software which was the Core development toolset which provided a win95/98 development ide for the sega saturn.. ;-)
Posted by nzamparello (60 comments )
Link Flag
They already re-use code.
Take a look at the internal workings of Homeworld and the follow-on Homeworld Cataclism. Take a look at Grand Theft Auto and any Need for Speed title. There is a reason these titles have the same look and feel, it is that the game engine underneath is being used. Heck look at EA's new Burnout 3 title and compare some of the vehicle rendering in it to EA's Need for Speed Underground. There is a good reason the cars all have the same kind of metalic sparkle.

I'm a huge fan of the Command and Conqure series, and C&C, C&C2, Red Alert, and RA2 all use the same game engine with ever updated graphics and new options pasted on.

Beyond same franchise we also see Unreal Tournament game engine being used in multiple titles by other developers. It is the engine used for the US Army game "America's Army Operations" for example. The biggest asset for the new Doom 3 is the engine and they are already seeking to license it out to other first-person-shooter titles for re-use.

What you are going to start to see is more cross-licensing of portions of titles. Rockstar games could profit from competing titles like Streets of LA and Driver 3 by licensing their game engine to those titles. They could also tap the resources of those developers for improvements to the game engine and incorporate in upcoming releases that would shorten the development life cycle.
Posted by (46 comments )
Link Flag
It just makes sense
I don't understand the logic in re-inventing the wheel every time a new release is rolled out. It makes sense to recycle ideas and code. Think about how many platform adventure games there were, now think about how many first person shooters. Driving games... etc. As long as things are constantly improving, why would it not make sense? There comes a point when you have to innovate to get things done. When graphics go up on an exponential scale, and the technology used to create the games stays the same, it looks to me like production will eventually come to a grinding hault.
They're not saying every game is going to be the same. It's like building a hotel along a one way train track that one train will ever go down. Once it's passed, do you go chop down all new wood?
Posted by (1 comment )
Reply Link Flag
You are missing the issue entirely!
I have been reading about this game developers woe issue for a year now. The whole industry needs a wake up call.

How long before they realize that the only way to develop games that are complex to point of insanity is to stop the one shot sale approach. Online games are not just profitable they are the future of the entertainment industry as a whole. No Artificial Intelligence will ever replace the feeling of human players competing and cooperating with each other, at least not in the next 20 years.

How can you afford to support bug fixes for a software application that has a billion lines of code? Online-based games that have monthly fees not only ensure that support is actually a profitable business model, but also allows new content to attract new game subscriptions.

How can you develop a sequel when every magazine and online review gives it bad marks when the graphics engine is not vastly superior? This reusable code concept isn't new by any measure of the word; it's been done since the inception of computer programming. The problem is that every game title has to be rendered consistent throughout, if you use a graphic from your WWII FPS in your Fantasy RPG you better believe that it will look a bit off.

The key thing here is that the online game will lengthen titles' lifespan from three months to three years. If you don't believe me check out an online game site and look at how many people are playing these titles, some of them are over three years old and have tens of thousands of people playing at any given time!
Posted by (2 comments )
Reply Link Flag
Now why did you go and have to tell them that? Like we dont already have enough small fry companies slapping together games with interfaces and physics 10 years old then selling them as online games.

The online game market is getting very cutthroat and players attention spans for the games are getting shorter. The moment they feel they arent getting their 14.99 a months worth they are gone when back in the EQ days they stuck it out through the borind times as well as the fun ones.

My advice is that if you intend to make an online game you better spend some time and money on it making sure at no point in the game playing experience is it going to be boring or frustrating, dont expect it to be an instant hit and under no circumstances should you EVER let a player violating the rules off with a warning in order to preserve your cash flow. Thats a death sentence.

Theres nothing more frustrating for a online gamer than to spend 49.99 on a game then $19.99 monthly for 2 months when you realize you hate the game and all you can do with it is use it as a coaster, you cant play it again one day if you get bored without significant hassle and you cant use memory editors to modify it for fun.

If they like it theyll love your company if they dont like it they will hate you there is no middle ground - beware.
Posted by Fray9 (547 comments )
Link Flag
This is just like any industry, survival of the fittest. Game devopment is going in a new direction and the old dinosaurs need to evolve or go extinct. In the other hand, the physics ARE an important creative aspect. Its the difference between that Thrasher game that came out about the time as THPS 1, one of them bombed and the other went to the top. Same thing for racing games, nobody ever said "Lets go play Ridge Racer Revoloution instead of Need for speed 3!". But, be real it takes way to long to make new stuff and I have to admit games, lately anyways, are REALLY lacking in intelegent gameplay with a few exceptions, KOTR was killer, I couldn't believe they put math problems in that game. Overall, I think this is going to be a good thing, a new era a new outlook on gaming if we're lucky.
Posted by (1 comment )
Reply Link Flag

Join the conversation

Add your comment

The posting of advertisements, profanity, or personal attacks is prohibited. Click here to review our Terms of Use.

What's Hot



RSS Feeds

Add headlines from CNET News to your homepage or feedreader.