Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
August 23, 2014
arrowPress Releases
August 23, 2014
PR Newswire
View All





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


Future is bright for AAA games, says Epic's Sweeney
Future is bright for AAA games, says Epic's Sweeney
November 13, 2012 | By Mathew Kumar




High-budget, high production games and technologies have a bright future, but the way these "triple-A" games are made in the years ahead must evolve.

That's according to Epic Games founder Tim Sweeney, the mastermind behind the widely-used Unreal Engine, speaking at the Gamasutra-attended opening keynote of the Montreal International Game Summit.

Sweeney described the lessons learned in the production of Epic's last two tech demos, beginning with 2011's Unreal Engine 3-based "Samaritan." He said Samaritan was a good proof-of-concept for the upcoming Unreal Engine 4, that pushed a lot of impressive, new UE3 features. But Sweeney said development of the demo was "greatly worrying," as it took a 30-person team and four months to produce.

"The result was just a three minute demo," he said. "We could see costs going up three, four or even five for the next generation, so it became apparent we needed to increase our tools for productivity so we could build our content very efficiently."

As a result Epic tripled the number of people working on tools in the process of building UE4, and produced another tech demo, "Elemental."

In discussing Elemental, Sweeney had high praise for DirectX11 as a way to improve engine efficiency, as it passes general computing processes through the GPU, allowing them to create effects such as real-time direct lighting.

"DirectX 11's compute features enable a huge leap in visual quality," Sweeney said. "The features are limited -- it took a lot of contortion to adapt our algorithms -- but sufficient ... over 50 percent of the GPU flops are in general computing algorithms rather than the standard pipeline, and we see that percentage only increasing."

Although not specific, Sweeney also said that Elemental also proved that their tools investment was paying off, with the demo developed at a "very rapid pace."

"But budgets are always going to continue marching upwards," he conceded. "We are hoping costs at the start of the next generation to only be double the cost of the start of the previous generation."

Mobile revolution

On the topic of "generations," however, Sweeney wished to make it clear he was not limiting that term to the console market, instead discussing the wider future trends and saying the increased accessibility of games via smartphones and tablets will be a "real revolution" in games, especially in developing markets.

"Smartphone and tablet performance is growing at an incredible rate," he said, arguing that it means "we'll be able to bring more and more of the value of triple-A gaming to these platforms."

Because of the increase in performance, Sweeney explained that it was now far easier to scale products to fit the platform, and that Epic is committed to building technologies that only increased that scalability.

"We think we can scale by a factor of 10-20 from low-end to high-end now," he said, "making it increasingly easy to ship across disparate platforms."

In fact it was in a discussion of these "disparate platforms" that Sweeney surprised himself, amusingly, with his own omission: reading from his slide, "There are three types of platforms. PC online, mobile, web-based... oh wow, I left out console! I mean no commentary by that!"

Of those, he considered web-based game development to be "really interesting," describing Epic's own partnership with Adobe to create a web-based version of Unreal cross-compiled into Flash. He felt it "wasn't quite practical yet" as a delivery mechanism for triple-A experiences, but over the next year improvements in cross-compiling technologies would make it "very feasible."

Triple-A free-to-play

Sweeney also touched on one of the biggest issues facing triple-A game developers today: the rise of free-to-play.

He argued, "In offering a consumer a free-to-play game via download or web, versus a $60 game available only in a store, the free-to-play game is going to win." He said that it was one of the "major interesting trends to watch as we go to the next console generation": whether the console manufacturers offer a significant uptake of free-to-play lessons and methods will "play a major factor in whether they are going to succeed or not."

Of course, Sweeney accepted that "hardcore" players wouldn't "tolerate that sort of model ... it's going to require a lot of development and iteration" to figure out ways to monetize console experiences without leaving those players feeling like "they're being ripped off or cheated."

Sweeney described Epic's future path as a developer, explaining that while at the start of the last generation they decided "console was going to win" and they considered their major investment to be partnering with Microsoft for Gears of War on Xbox 360.

In the next generation, however, "we are not focusing on any one thing," he said.

"[Epic will] build some games for PC, some games for mobile, and some for console," said Sweeney. "And any time we make something for one, we're going to explore how it fits on other platforms ... we are going to be building a lot of multi-platform experiences."


Related Jobs

Raven Software / Activision
Raven Software / Activision — Madison, Wisconsin, United States
[08.23.14]

Sr. Software Engineer (Gameplay)
FarSight Studios
FarSight Studios — Big Bear Lake, California, United States
[08.22.14]

Lead Android Engineer
Churchill Navigation
Churchill Navigation — Boulder, Colorado, United States
[08.22.14]

3D Application Programmer
Glu Mobile
Glu Mobile — Bellevue, Washington, United States
[08.22.14]

Lead Engineer










Comments


Alex Boccia
profile image
I like Sweeney, he's a pretty cool dude. That goes for Epic in general, too.

Joe McGinn
profile image
Agree with most of what he said, except that hardcore gamers have trouble with free-to-play. League of Legends is a hardcore game no matter how you slice it.

Terry Matthes
profile image
Yes. I'm interested to see what Epic Games does now that Cliffy B is gone.

Raymond Ortgiesen
profile image
Future is bright for candle making, says candle maker.

Tom Baird
profile image
And the Best Selling Mario game came out in 1985.

Are you equating AAA games as a whole to the fact that a single console-exclusive franchise managed to outsell it's previous predecessors, when using significantly more resources to do so?

http://www.joystiq.com/2012/10/25/spencer-halo-4-is-microsofts-mo
st-expensive-game/

If Halo 4 doesn't massively outsell it's predecessors it is still a worse ROI than previous iterations. I would argue that inflation of dev costs to stay in the AAA space is more harmful to AAA health than simply declining sales.

Johnathon Swift
profile image
No no, huge triple A games are always going to be around, they'll be huge. Just try thinking long term. "Do I want my own Holodeck?" The answer is obviously yes. Huge virtualized worlds are going to keep getting bigger forever.

But the question of how and when we get there is rather blurry at the moment. "Free to play is the winning solution!" Because why? League of Legends, one game? People are stabbing at everything. Guess we'll just wait to see what sticks.

Luis Guimaraes
profile image
Future is dark, says another candle maker.

Vau Dev
profile image
"Future is dark, says another candle maker."

Obviously, he doesnt know how to make candles.

Luis Guimaraes
profile image
@Vau Dev

I'm not sure if you didn't get my comment of if I didn't get yours.

Jesse Tucker
profile image
"But Sweeney said development of the demo was "greatly worrying," as it took a 30-person team and four months to produce."
As much as I was impressed by the Samaritan demo, this was the same reaction I had. As a level designer, I'm really excited to be able to make really outstanding areas. However, the thought of working on a single area in painstaking detail for months doesn't seem too appealing. The ability to drop a light in an area and have it "just work" sounds fantastic, and is a great step in the direction of letting people make excellent-looking work quickly.

Andrew Traviss
profile image
"Only" doubled? So the number of sales has to be twice as large or the price to the end user has to be twice as large in order to achieve the same margins? This is going to go well.

Michael Pianta
profile image
Seriously, that's not good at all. I suppose we can expect the frequency of all the "Studio X is shutting down" stories to double as well.

Darcy Nelson
profile image
I almost spit out my drink when I read that line. I wasn't sure if the title of the article was meant to be ironic, after that.

MaurŪcio Gomes
profile image
I wish epic relied less on DX... Since that means pushing farther and farther on windows version, and new windows versions are not THAAAAAT good.

But let's wait and see.

Ken Kinnison
profile image
Wow I agree on something vaguely anti MS... considering how close to an MS fanboi I am this makes me feel slightly ill...

Eric McVinney
profile image
Future is bright? Really? How many times have we heard this? I'm sure that it's not a matter of WHO, but of what product that will be released on what platform.

Jorge Molinari
profile image
Why do game engine companies keep focusing so much on graphics? Clearly, what is preventing video games to evolve is the lack of realistic physics and realistic AI. Without these two elements , the next generation AAA games can only aspire to offer the same experiences of the current gen, but with a shinier coat of paint. Is a shinier coat of paint worth the increased risk that these new hyper budgets will add to an already very risky industry?

Of the two missing elements, a realistic AI is the much tougher nut to crack and will probably take decades to solve. That is still mostly the domain of University and Government research, not game developers. But why do we still have such crappy physics in 2012? Isnít it all just equations? You know, like lighting, particles, and meshes? Why does every increase in computational power have to be dedicated almost exclusively to improve graphics while the physics in games remain almost as crappy as 15 years ago? Why do the engine creators continue to push polygons instead of physics? Is working on physics that much harder? Do they take that much more processing power compared to these new fancy effects of the latest 3D engines? Do they think that most consumers prefer graphics over physics? Personally I would be very happy if the graphics donít progress for an entire decade but we get real physics into games.

Grand Theft Auto 3 and Grand Theft Auto 4 are separated by a console generation. Yet it is still essentially the same experience, only GTA4 has a shinier coat of paint. Is anyone here excited about the next GTA? How about Halo 4? Am I the only one that is tired of the same formula? (Package the same experience where the only change is a new story arc, with incremental improvements to graphics, and a pinch of change to gameplay? Not that I donít appreciate good story and graphics, but letís be honest: even a crappy movie will feature better story/script and visual realism/artistic design then the best video game.

If any AAA engine developer is reading this please take away this message:

Many of us donít want a game that LOOKS less gamey. We want a game that IS less gamey. Please stop developing graphics and bring the physics up to par.

Eric Geer
profile image
I'm not a game developer, but I agree with this entirely. One of my favorite games this year was Rage. Although many people have discounted the game as a "tech demo" I found this game drew me more into it's world than many games, not because it looked fantastic(looked damn good) or because of the story, but rather the AI. The way the enemy AI moved and reacted seemed more natural than most any other game I have played, yes there were some hiccups, but I would say it moved in more the right direction than many games have even tried.

Mathieu MarquisBolduc
profile image
There is a dissonance between what people say and do. People say they want smoother gameplay, smart AI and realistic physics, but overall shiny graphic is still the #1 seller.

"even a crappy movie will feature better story/script and visual realism/artistic design then the best video game."

This I could not disagree more.

Joe Wreschnig
profile image
"Why [do] the physics in games remain almost as crappy as 15 years ago? Why do the engine creators continue to push polygons instead of physics?"

This is really utterly false.

15 years ago - and let's be explicit, that's 1997, 1998! Quake 2 was new! - only a handful of high-end games had anything resembling a physics system, and then they usually just did boxes or spheres. Most collision systems and character controllers were ad-hoc for the game's requirements.

Today I can download one of a half dozen free rigid body physics engines for 2D or 3D, with all kinds of simulated forces including friction and angular momentum, all kinds of possible joints and constraints, arbitrary meshes, etc. Although the tooling isn't as ubiquitous, plenty of games at all levels of budget have soft-body engines, for liquid and cloth simulation. Some games use bespoke systems to attenuate light and sound based on physics properties of materials. I can synchronize physics simulations over the Internet, which barely existed in games at all 15 years ago!

15 years ago I was lucky if I could walk across my ship mesh in Daggerfall without falling through. Today the trademark of Skyrim is a physics mechanic that sends dozens of objects of disparate size and AI flying through a complicated landscape, and most people don't even consider it twice.

Michael Joseph
profile image
People say they want to be in great shape yet they don't work out and eat bad food.

That inconsistency doesn't mean that they actually want to be fat. It just means that people have frailities.

Duong Nguyen
profile image
Very few studios really have the resources to do these types of "cutting" edge research anymore, they mostly use 3rd party engines like UE4..

We have rigid body physics in games, they are close approximation to real world but obviously that's no fun so u have "game physics" with magical forces where needed. You can spend the extra cycles on fluid dynamics and soft body, but unless the game really needs that ( some games do ), it doesn't offer the return warrant by the investment of time and energy.

AI is such an open field, games are not going to solve the fundamental issues of intelligent AI, at best they can create reasonable AI for their worlds. There are a few 3rd party libraries which can help solve some of the low level issues like pathing but the higher level behaviors are still very much context dependent upon the game. Solutions develop over time and incrementally improved. That's why u don't see companies just jump into a new genre and have great AI, it usually takes 2-3 iterations before it gets acceptable that is if they even stick around that long.

Patrick Lavelle
profile image
This may be simplifying it, but it's easiest for the consumer to judge quality based on the visuals, brand, and features of a game than to experience several hours of it and then make a purchasing decision (if they can even play before buying, which is the strength of demos, shareware, and F2P). The publisher's marketing department knows this, and so the developer is encouraged to focus on the things that will best sell the game. It's peacocking.

I think the problem is the saturation of content and the difficulty of filtering out the best on the digital market, in addition to the classic publishing problem where they are unable to identify or incubate both high quality and progressive games.

David Canela
profile image
Physics and AI are great, but definitely not a case of "more is always better". Many a game has artificially dumbed down enemies because in some cases that's actually more fun. It's the same with everything, really. More realistic graphics aren't necessarily better graphics and the same applies to physics and AI.
At the end of the day, you need what's right for that very specific game. If you're making some sort of simulation game, yes, that means less gamey physics etc. But in other cases, more gamey is just what you need (it sounds weird to me to complain about games being gamey, books booky and films filmy...)

Crimson Ghost
profile image
I don't mean to be snarky, but the idea that AAA development needs to evolve should be clear to anyone with experience in the business.

Also, Epic's technology itself could also stand to be improved. I know designers have enjoyed working with some of the tools (though some haven't) but I know many programmers who hate it.

Ian Richard
profile image
30 dev's and 4 months for a 3 minute demo.

Call me crazy... but that doesn't look bright to me.

Duong Nguyen
profile image
They are getting better I hear their new demo "Elemental" only took a small team a few weeks, but it's no where near as impressive as the "Samaritan" demo imo.

Patrick Lavelle
profile image
I would be curious to know what resources and time were required for the UE3 demos.

As those of us who have done demos know, they are very detailed and specialized and are not representative of the cost of content for the actual game.

Chris Remo
profile image
"We are hoping costs at the start of the next generation to only be double the cost of the start of the previous generation."

How is this even remotely sustainable in a world in which studios are already closing at an ever-increasing rate, unable to survive in the current triple-A world of budgets rising without a commensurate rise in the size of the $60 game-buying audience?

Johnathon Swift
profile image
Borderlands 2, Assassin's Creed 3, Halo 4, GTA V. Ok I don't know why it worked out in sync like that. But all those can afford more money, or have a big enough budget that they don't need to increase it in the case of Halo and GTA.

But no, it's not sustainable for most games, at least not in the current publisher/developer business model. Which is why Kickstarter and Unity are doing so well. Games that cost less, who knew?

But heck, maybe games will be able to double in budget. Hollywood makes $200 million movies and get's away with charging $10 a pop for huge profits multiple times a year. Why can't video games?

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Johnathon Swift
profile image
"Movies are much more mainstream than games can ever hope to be..."

No. Just, no. More accessible? Yes. Because they only require eyes and ears and no input they can lower the bar to least common denominator willing to pay. Therefore it's EASIER to get a movie that more people are willing to pay for than it is to get a game that equals such. But games have a much wider possibility of existence. You could easily spend a billion dollars on a game, and once we have the technology a billion dollar game will certainly blow past the total for any movie ever made.

We're just not there yet. But the point is we will be. Humanity does not stop advancing because we hit some minor snag. In fifty years we'll have that billion dollar game, and there will be a couple hundred million people playing it. How we get there is going to be interesting though.

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Jorge Molinari
profile image
@ Joe Wreschnig

Iím not going to count ďnot falling through the meshĒ as a physics advance. Collision detection may technically be physics but címon now, this is a pre-requisite to even consider making a game. Iím glad you mentioned Skyrim because that game comes immediately to mind when I say I would like the physics to be on par with the graphics. I consider Skyrim the gold standard in immersing the player in a virtual world. And yet the AI and physics feel generations behind the rest of the game. (The bow/arrow physics are pretty darn good though.)

So yes, from a developer perspective you now have some off-the-shelf tools to make physics simulations; but as an end user I havenít seen much of that translate into actual AAA immersive titles. Here are the physics advances Iíve noticed in nearly all AAA titles:
* Cloth that is not flat and can move. (Still, in most games cloth in the environment does not drape realistically over your avatar.)
* Water that is not flat. (Iíve never seen it flow)
* Ragdolls. (But the dead bodies become nearly weightless and frictionless as my avatar pushes them around on the floor)
* Explosions that make a dent in the ground mesh.
* Glass windows break. (Not really based on real physics anyway)

And here is some physics advances that feature in a very short list of AAA games:
* Destructible walls. (They are discrete areas and completely unrealistic.)
* Collapsing buildings, and I mean based on physics. (Silent Storm is the only game I know of that gave it an honest try.)
* Destructible world mesh; I think? (Itís been too long, but I think the original Red Faction allowed you to make a hole in the mesh wherever you pleased, Iíve never played any of the sequels)
* Objects in the environment that can be picked up. (Besethda RPGís, works more like telekinesis. Admittedly I canít think of an interface to make this work.)
* Debris from exploding objects are actual projectiles that fly and bounce off the environment.

Canít really think of anything else. So from my gamer perspective, the universal advances (first list) have been few and in between, and are still not realistic anyway. I think there is a lot of game to be had if someone bold enough would use todayís computational power to make a world that behaves more realistically, even this means using graphics from 5 years ago.

David Canela
profile image
Realistic (and thus complex) behaviour may be great for some games, but can become too sandboxy for other types of experience and make it even harder for designers to have even a rough idea of what will happen when players play. Not saying that's a bad thing, just saying design considerations might play just as large a role in why you don't see complex AI and physics that ofen as actual technical challenges...

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Terry Matthes
profile image
Hi Jorge I agree it would have been nice to see some better physics in Skyrim. I think it might have been a resource issue on consoles given the high quality of all the other aspects of the game. Perhaps it wasn't worth the effort as PC players might have been the only ones to get the benefit; as I remember PC was only 10% of sales.

*edit

I found some sales numbers dating back to November. These numbers are only physical sales. Digital downloads aren't included, which I admit would skew the numbers in the favour of consoles

Xbox 360: 53.1% (2,745,921)
PS3: 28.5% (1,475,466)
PC: 18.4% (949,536)

Mr Developer
profile image
This is just the same old sales pitch from Tim Sweeney. Epic's only tech strategy is to dazzle naive developers with flashy visual effects that distract them from that fact that they are buying the same tired old code base and tools.

Anyone with experience using their product knows that every team ends up rewriting most of their code base and fixing all their bugs in order to ship.

Experienced developers are pushing boundaries Epic can only dream of such as the living worlds Rockstar excel at or the technical excellence and polish Naughtydog consistently deliver.

I sincerely hope UE4 will provide something of substance for developers to improve their efficiency but my experience tells me that is very optimistic.

Ramin Shokrizade
profile image
The microtransaction bubble pulled 87% of all investment resources last year away from AAA. This leaves the field largely without competition for those AAA games that do make it to market in this environment. Presumably, with the bubble now bursting, resources will move back to AAA. Nonetheless, these titles will be under intense pressure to adopt F2P monetization models. This does not necessarily mean microtransactions, however. It certainly does not require "pay to win".

I should also point out that unless you have a clear reason for making your game multiplayer, and have built it that way, then don't. This was one of the problems with Star Wars:TOR as I pointed out in an earlier article.

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Nooh Ha
profile image
"The microtransaction bubble pulled 87% of all investment resources last year away from AAA"

This is a very odd statement.

Firstly, by saying "away from AAA" you appear to be suggesting that all that investment would have gone into AAA games development if it had not gone into microtransaction gaming. You obviously dont get to deal with VCs or private equity companies much :> Traditional console and PC retail games have been a no-go area for most VCs for at least 5 years if not longer.

Secondly, VC investment (which I assume your 87% figure refers to) represents only one source of games development funding. Publisher funding remains not only the principle source of financing for retail games but, in aggregate, comfortably exceeds VC investment by several multiples. Just have a read through the SEC filings for the top publishers...

Michael Joseph
profile image
Natural Motion's (Euphoria folks) clumsy ninja demo...

http://www.youtube.com/watch?v=liLzcdc984A

this is just one example of the types of systems modeling that will demand greater processing power to run but which can help drive the industry and drive hardware sales.

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Michael Joseph
profile image
"I guess what I'm saying is I don't see the dynamics adding more than musing to the believability of the character"

People used to say similar things about Quake's polygonal character models over the sprites which had been used up until then.

I think what one could do with AI/procedurally driven character behaviors and animations is significant enough that once one successful game starts doing it, everyone else will follow.

p.s. I suspect Natural Motion is filing alot of patents around this type of tech. So while the industry at large continues to grab for the rapidly diminishing low hanging fruit, there are others out there who are really looking towards the future and working to define it. And it's not Epic and UE4 by the sounds of it.

[User Banned]
profile image
This user violated Gamasutraís Comment Guidelines and has been banned.

Terry Matthes
profile image
I think a lot of people forget that 100% of a game's budget doesn't go to development. In fact it's probably quite a bit less when you consider marketing. I only mention this because I think it's terrible most people would just think of slashing the dev. budget as opposed to the marketing budget to make up for these ballooning costs.

If a developer had more money to work with on their games one would hope they could produce a better product and wouldn't have to spend millions to try and convince everyone on a superficial level that their game is worth playing. I say "superficial" because most of the ads coming out for games like Need for Speed and COD don't even highlight the gameplay. It's based around making you feel like you (anyone) could be an elite member of that niche group whether it's the army or a secret racing club. Campaigns like these literally piss millions of dollars into the wind hoping to grow their user base or should I correct myself; sales numbers?

Sorry but this stuff makes me angry so I'm going to continue... Advertising like this will get people to try your game, but it's not a organic way of growing your user base. Most of these people aren't going to be converted into long term players anyway. But the sales numbers look good so that must mean the franchise is really healthy right? If the core players feel the game isn't up to snuff you're not going to have a passionate community at the core of your game and that's what keeps a franchise going. The tactics these ads use are disingenuous at best and akin to lipstick on a pig.

I got a little of topic there, but I feel a lot better after getting that out. Thank you Internet :)


none
 
Comment: