Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
November 1, 2014
arrowPress Releases
November 1, 2014
PR Newswire
View All





If you enjoy reading this site, you might also want to check out these UBM Tech sites:


We won't make the same mistakes next generation, says EA
August 30, 2012 | By Mike Rose

August 30, 2012 | By Mike Rose
Comments
    33 comments
More: Console/PC, Business/Marketing, Video



"Frankly, we made a couple of mistakes in how we built our technology that was going to build all these games."
- Frank Gibeau, head of labels at Electronic Arts admits that the company had a rather rocky start at the beginning of the current console generation.

However, Gibeau says that he has pinned the problems down to one point in particular - technology. The company didn't have a powerful enough game engine in place to properly take its games to the necessary level, he says in the Bloomberg interview. With its proprietary Frostbite engine, EA is making sure this doesn't happen again.

"[Frostbite] is tailor made for next-gen hardware, so we're in really good shape from a technology standpoint, which is where we had our misstep last time," he notes.

A rich and broad set of intellectual properties is also something EA was missing at the start of the current console cycle, says Gibeau, and again, this is something that the company is looking to rectify come the next generation.

"We've already started three to five IPs that we're going to launch in those first 24 months of the next generation," he says. "EA Sports is there with all of its power, and you also see some really big brands like Battlefield coming out."


Related Jobs

Twisted Pixel Games
Twisted Pixel Games — Austin, Texas, United States
[10.31.14]

Senior Graphics and Systems Engineer
Twisted Pixel Games
Twisted Pixel Games — Austin, Texas, United States
[10.31.14]

Mid-level Tools and Systems Engineer
Giant Sparrow
Giant Sparrow — Playa Vista, California, United States
[10.31.14]

Junior 3D Artist
Giant Sparrow
Giant Sparrow — Playa Vista, California, United States
[10.31.14]

Lead Artist










Comments


Rob Wright
profile image
Um....does ANYONE here think EA's biggest problem was graphics? Hello? Crysis? What about Mass Effect or Dead Space or freakin Medal of Honor for that matter? I have a hard time accepting that EA's problem was visual in nature. Did Spore underperform because it didn't have a next-gen graphics engine? Come on, seriously....

Raymond Ortgiesen
profile image
It's true that he might be referring more to the in house toolsets they're using. Having an incredibly inefficient back end is going to slow down iteration on those titles. You're essentially right though, to say that all of EA's problems boil down to technology is disingenuous of Gibeau.

Eric Peters
profile image
Did you read the story at all? He was talking about the graphics and the whole engine behind them in their current gen launch titles and the titles that followed closely after. And yes they did have a big problem with graphics. Most of their titles at the start of this gen were nothing more than high res versions of Xbox 1 games. They bumped up the poly counts by a decent amount, but that's it. There was nothing that really "wowed" anyone at launch, nor in the months or even the first year or two that followed. None of the games you mention came out in that time frame, and in fact, where 2nd or 3rd gen games, and in Crysis's case, it's not even a EA game, it's a 3rd party title that was only PUBLISHED by EA......

There is no doubt EA was slow out of the gate visually this gen, anyone has to do is compare screen shots of it's games to other devs at the time, or hell to xbox 1 games to see this. But that of course was not their only problem, nor their biggest. At others have stated, their biggest issues were from gameplay and innovation, not in graphics. The fact that the CEO is listing graphics and underpowered engines as one of their main "mistakes" from this gen only serves to show how out of touch they are as a company and how little seems will change with the next-gen as far as EA goes...

Rob Wright
profile image
@Anthony

True, but unless he's talking about EA games for the Wii, I don't see how he can claim with a straight face that EA stumbled at the start of this cycle in 2006/2007 because of lackluster game engines. EA's problem is a content problem. What big, critically acclaimed franchises besides Madden and Sims did they have to bet on back then? They didn't get their Medal of Honor/Battlefield franchises in order until Call of Duty exploded with MW. Plus, they made some bad bets on new IPs (I'm looking at you, Hellgate London). Again, I don't think it's credible to claim a proprietary game engine would have reverse that.

Rob Wright
profile image
@Eric
1. Yes, I read it, and watched the whole video interview (the PopCap stuff was a little uncomfortable).
2. You are correct about the graphics quality of those launch games, but I'd argue that was true of the vast majority of games launched in those first few months. This was especially true of the Xbox 360 which was rushed a bit to beat Sony and Nintendo, thus developers didn't have a heckuva lot of time to develop graphics that were much more than glorified previous gen visuals.
3. True, the launch time frame is much narrower than the first 2-3 years, but again, as we've seen with previous console launches, it usually takes a year or two to really start maximizing the graphics cababilities. There are exceptions; Gears of War was released in the first year for Xbox 360, and it's pretty stunning. But overall, I think it's a little crazy to expect a huge upgrade in visual performance right out of the gate. Just my opinion, but you get a better sense of how a pub/dev is utilizing the console hardware after 18-24 months, and a slow start certainly isn't going to ruin the entire console cycle for you either.
4. Well, Bioware was acquired by EA before Mass Effect was released, and Dead Space was developed by EA Redwood Shores (now Visceral). I see your point about 2nd/3rd party games. But what does that really matter? Why would EA develop a top-tier proprietary game engine like Frostbite and then NOT license the engine to third party developers for EA published titles? That wouldn't make much sense, would it?

Bob Johnson
profile image
@Rob

EA also has Need for Speed and Fifa. FIFA sells more copies than Madden. And you may be surprised at how big Need for Speed still is or was. It has declined this generation although the Criterion-made 2010 NFS:HOt Pursuit sold 8.5 million copies according to Wikipedia. Again more than Madden sells in a year.

I do think EA had a slow start. I mean didn't they run their basketball game into the ground? Their early Madden's were criticized fairly heavily for bad graphics. And yeah I remember the HD-enhanced versions of Xbox games during the launch window for full price or close to it.

I think EA put its bet on PS3 development up front too but that platform proved to be the slowest starting platform and the 3rd place platform in the US. And so that hurt them.

And overall EA just releases a shoddy product. They are really big on hitting their back of the box bullet point features and probably should be concentrating on fewer but better options/modes in their games. And actually working to make a better, more refined product.

Michael Rooney
profile image
I definitely don't think he's talking about graphics in particular. Working on a wide range of iterative titles, it's very important to have your main pieces of technology orthogonal to your game code. Otherwise game teams end up reimplementing similar features and wasting a lot of time solving the same bugs or having two branches of the same core technology that are totally incompatible with one another.

You have to remember that users see the final product. The product might look fine to you, but it's more than likely in spite of workflow/development inefficiencies not because there is a lack of them; this is true of pretty much all software engineering. If those inefficiencies are significant, just removing them will dramatically influence the quality of the end product.

E McNeill
profile image
New IP is always welcome!

Kale Menges
profile image
Just a quick note: Technology doesn't build games; game developers (you know, PEOPLE) build games.

Mathieu MarquisBolduc
profile image
No, compilers & linkers build games ;)

Michael Rooney
profile image
Give a carpenter a hammer and see what he's able to build in a week. Give that same carpenter a fully furnished workshop and see what he is able to do in the same time.

Technology doesn't necessarily affect talent or knowledge directly, but it does have a significant impact on how well people can take advantage of their talent/knowledge.

Andrew Grapsas
profile image
Having worked at EA, and without getting into specifics, I can say that this actually does make sense. I do think EA is, currently, in a much better place for entry into next gen than the previous gen.

While tech doesn't dictate game design, it certainly paces development, which is a huge factor in the quality, delivery, and details of the end result.

Mathieu MarquisBolduc
profile image
I concur. People might not remember that at the turn of this generation, EA had just purshased Renderware (through Criterion) as their new engine of choice, which, as EA themselves stated, didnt work well at all on the new gen. Plus Renderware was mainly a graphic engine, where the competing engines included tools for anim, gameplay, ai, level design, etc.

Nathan Zufelt
profile image
Agreed. I didn't hear anything about "graphics" like the posters above are frothing over.

Not having every single team in your company using their own proprietary engine and toolset is a huge advantage when it comes to sharing resources and knowledge.

Of course "technology doesn't build games, developers build games".... I didn't hear Frank say anything to the opposite of that. What he said was that the company ran into tech hurdles with the last console transition and that he believes the the company is in a much better position to handle tech issues with the launch of the next console generation.

Michael Rooney
profile image
I don't really like my comment so removing it.

Andrew Grapsas
profile image
Triple-A studios are realizing that content and assets are becoming huge bottlenecks and expense points for development. So, build better tools! Unreal is a great example. Unreal 4's tools are lightyears ahead of Unreal 3's. IMO, the reasons are simple: as we can push more pixels, it's less about building the technology of pushing pixels (that's very well understood) and more about increasing productivity.

When I started in this industry, I had to wait 30 minutes for a PS3 build. The last triple-A game I worked on had that under 3 minutes. I'm sure other studios have done an even better job. Now, on web and mobile, I hit compile and it's basically done.

With on-the-fly compilation of C++ (as was demo'd in Unreal 4), gameplay programmers will have a huge advantage! Not to mention how much more streamlined UnrealEd is.

I have no doubt in my mind that Frostbite is doing the same. As Frostbite becomes more pervasive throughout EA (I've never worked with the engine, and haven't asked any friends what their experiences are, so, guessing here), the tools will become better and better.

I've worked on projects where tools have certainly held back the project. Obviously, good game design, great people, etc. all are vital; but, with bad tools, you, more than likely, will still get a bad game -- obviously there are exceptions, as Uncharted's first iteration was built with barely any tools on unproven technology (Source: http://www.gameenginebook.com/)

Kelly Kleider
profile image
Renderware had tools for animation/level editing etc...

EA wanted their own middleware and they also wanted to deprive their competitors of the same. GTA3+ were built using Renderware. Rockstar didn't use a whole lot of the "extras" that Renderware offered.

If EA had done a proper evaluation, they would have realized Renderware was not a robust solution for them and either rolled their own, or found another product. Good graphics engines are plentiful, it comes down to solid tools (art, design, audio) and ease of use (engineering). Without those pieces solved they will suffer a similar fate for the next launch cycle.

Kelly Kleider
profile image
@Andrew /to Andrew's point about good vs. bad tools.

Good tools are value neutral, but bad tools have a measurable negative effect. In other words, good tools won't make you a better developer (that's on you), but crappy tools can make a good developer produce bad results.

Michael Rooney
profile image
@Kelly: If by "good tools" you mean tools that can not be improved in any way to help the developer work faster, then you are correct.

I do not, however, know of any tool regardless of it's goodness/badness that cannot be improved in some way. For example, Visual C++ is a good tool, Visual Studio is a better tool, Visual Assist makes Visual Studio better still. Would you not say a good developer would be more productive with VAX than VC++? If they both add no value, then VAX shouldn't even exist.

Kelly Kleider
profile image
@Michael
I didn't say anything about productivity. Do any of those tools you mentioned make you a better programmer?

Mathieu MarquisBolduc
profile image
You said "good tools are value neutral". Are you saying productivity has no value? Better tools help me code better and faster and help me make less mistakes and help me find and fix those mistakes faster. So yes, those tools make us better programmers. They will not make an expert out of a beginner but they make us work better. Tools are everything in AAA production.

Aaron Casillas
profile image
Being flagship is key on a new tech platforms, this sounds like an honest critique of this generation. Sounds like a plan.

Mikhail Mukin
profile image
Yes, glad EA finally sees one of the big problems! Almost each EA company was building their own engine/toolchain (some companies - like one I worked for - were building like 4+ of them at the same time). Mostly duplicating functionality. It is by far not only graphics... It means that out of 3 years dev cycle, first year design can't do pretty much anything apart from paper design (and web surfing:). Second year they suffer with first versions of broken tools, unimplemented functionality, various things that needs to be worked around (and just plainly not knowing what they can and can't or how to do it yet). Beginning of the 3rd year, things begin to solidify and at the last half a year or so maybe we have something close to working. And this is when the game is sometimes almost re-done (design wise). As my project director was joking about our PS2/XBox1 title "we re-made the game in the last 3 months", and with our first PS3/X360 I'm not even sure if we (people responsible for tech and tech production) gave them even that much...

After in house engine was written and a game shipped, in the middle of developing the next game, Frostbite was "dropped" on us. No documentation. No support. Can not talk or call anybody on the frostbite side. Frostbite C++ (it was ~2+ year ago, hope it changed!) had some big files with not a line of comments in them... Our "evaluation team" (I was part of it) spend 3-4 weeks just trying to figure out what this engine can do, what the limitations are, can it work for big world streaming game with dense worlds and a lot of combat with all sorts of vehicles by splitting engine into systems and trying to figure out how they worked... At the end - the results were IMHO inconclusive (Frostbite has a lot of good things, as our engine, that we built during last 3.5 years, was already doing a lot of things too). Weeks of evaluation work for the senior tech/art people that could have been replaced by a few hours of Skype chat with members of frostbite team!

Obviously, another thing developing in house engine does is raising budgets to something close to $40 mil for AAA game, making most of them unprofitable... you need 25 engineers to make this happen, and your army of designers and artists is really not doing that much for the first half the project (but as our industry is still mostly non contract - so you have all those people on payroll...). Plus, long QA time for new tech adds more $. As a tech producer who actually likes to counts things :) I would say a game like ours can be done within $15 mil range, with a good starting tech (say, Frostbite), with the right people, properly motivated, aggressive use of outsourcing and contractors, etc...

Moving to one flagship engine is good. Some parts of EA central tech are very good. But there are ways to do it right and do it wrong. In the last year before EA close our company, it was almost painful to watch (helplessly) how lot's of money are wasted due to improper organization... Looks like EA approach to organizing work was (in many cases): "so you can't program... you can't design... you can't make textures or models or animate them... but you can do high level game presentations... so maybe you can manage the development then?". While I understand that people who can present the game on a very high level are very important... people who can actually manage the development process are just... different people!

And for all 200 (or I don't know how many they have? VPs)... How about maybe visiting a studio like once in a half a year and talking to people? I mean, people who make games, not to some senior producer, who understands the game mostly from a series of Excel tasks/resources spreadsheets (w/o knowing what many of those lines are and why they are there)?

As EA shareholder... those VPs/etc are responsible for loosing thousands of my money in the recent years, so let's say I'm not very happy not only about years of my life... Well, on the positive side... those shares were mostly granted by EA anyway, so I guess I should not complain! ;)

But I'm sure EA can do better, just too much talent in various EA studios! So EA, I'm keeping my shares for a few more years and expect results!

Aaron Casillas
profile image
Yes sir, big company game development needs a review...

Bart Stewart
profile image
The question, though, is what is the most important strategic action for launching on the new boxes with compelling games.

Having a new engine ready to take advantage of new hardware seems smart. But is that really more important than changing the top-down, risk-averse culture that produces bullet-point games with big marketing budgets?

I'm inclined to think the most productive strategic action EA could take would be to get the suits out of the design process.

Nathan Zufelt
profile image
It can be really difficult to innovate and play with new ideas when struggling to get the baseline tools which allow you to create the "simple" stuff.

David Serrano
profile image
Agreed.

A W
profile image
From a consumer stand point (mine) EA's problem is in the "Me Too" marketing of their games as of late. They have had multiple IP chances that they just dropped the ball on or didn't let grow past the first (or second in some cases) inning. They produce games at first and perhaps though collectively to themselves "Yeah, we are EA and we are synonymous with Quality" but it just didn't pan out that way in the consumers mind. When EA saw that the others developers move towards specialization and began to surpass them in the market share of games which they had mastered designing on older systems (Medal of Honor as an example) then they moved to a "Me Too" strategy. They started trying to prove how their game's game play like; Dante's Inferno for example, is like that of God of War or some how better than it. Then when that strategy didn't work out they moved to the "premier graphics endorsement" and started marginalizing the console gamers for the computer gamer with the tag line "best on PC" because of the graphics. That just translated in to "we spent more time optimizing the high PC settings and then cutting out the fat to get it to run on the lower systems". Not a good method of designing and not a good method for selling to the AAA circles that exist on lower end tech IMO.

So from my standpoint non of those designs worked well for them in acquiring larger sums of money and now they are moving towards their next theory. This time I hope its more inclusive than exclusive and less "me too" and more this is the direction.

Justin West
profile image
@Rob, EA stated they needed to work on their own proprietary engine, which is Frostbite. Crytek owns Cryengine, Mass Effect was done with Unreal, and DeadSpace with Havok. This visual nature was done with other people's engines... So yes, they were behind like he stated, on their own technology stand point.

John Flush
profile image
That's amazing news. Now when will you get rid of "Project $10"? Forced social to even play? turning off content mid-generation? Sounds good for devs and all, but I still won't spend any money on EA anything. The logo is all I have to see these days to look elsewhere.

Chris Jimison
profile image
To me the thing I find very interesting is EA always touts how they are number 1 (in this case mobile), yet their stock value has seen nothing but a continuous decrease. This means 1 of 3 things to me:

1) They are lying (which I don't believe)
2) They don't know what they are talking about and are judging the market in metrics that no longer apply (what I do believe)
3) Their cost are so high that even being number one can not turn them a profit (what my gut tells me)

I worked at EA during the last console transition and I truly hope they do not repeat the same mistakes but I see it happening again. Frostbite maybe a great piece of tech, however the reason people license Unreal is because of the tools, documentation and general support which takes years to buildup.

David Serrano
profile image
It seems like many of the large developers and publishers are blind to the fact that millions of players didn't leave the AAA market over the past four to five years because the graphics were sub-par or the hardware was underpowered. Long time players left and new players stopped entering the market because of the extremely narrow, derivative, inaccessible and misaligned focus of AAA development. Combined with the increasingly exploitative and manipulative business practices of companies like EA and Activision. So if execs like Gibeau believe that higher definition versions of derivative games running on more powerful hardware will solve their loss of market share problems, they are sorely mistaken.

As was the case with the 360 and PS3, there will be a 12 to 18 month new toy bubble effect following the launch of the next gen. consoles. After this period has ended, if AAA developers and publishers haven't shifted their focus away from primarily serving the 16 to 23 year old male hardcore shooter - multiplayer audience and towards primarily serving the larger, more demographically diverse audience of adults... they'll be faced with the same market share problems they already have. But with seven or more years left in the hardware's life cycle. So for the next gen consoles to succeed, the industry must start to radically reinvent itself now. Which pretty much demands removing people like Riccitiello, Gibeau and Moore from the business and development equations.

Mike Kasprzak
profile image
There's going to be a next generation?!?


none
 
Comment: