AAA games are cheaper to make than you think

[MetalCraze]

http://www.warhorsestudios.cz/index.php?page=blog&entry=blog_007

You just need to have people who play them developing them.

 

[Infinitron]

Cool story bro

 

[Wirdschowerdn]

That was actually a pretty good read, and Daniel Vavra is a true bro. I'll keep checking what his new studio is coming up in the future.

It's a CryEngine 3 RPG right?

 

[Destroid]

High quality indie productions with small teams like Natural Selection 2, Overgrowth and Hawken support this article.

 

[deuxhero]

Even if AAA games got a lower budget, the money would be put to marketing instead.

 

[MaroonSkein]

Any hobo can make a AAA game. It's all about AAAA games now.

 

[Chuck Norris]

It's always a good idea to summarize what long ass articles are about ( or at least quote the important parts ) when you link to them. Some of us are busy here. I can't and don't want to read 157 lines of text about how to make AAA games cheap. I like to know about it, I'm interested. I just don't want to read the whole fucking article. Thanks.

 

[deuxhero]

It's always a good idea to summarize what long ass articles are about ( or at least quote the important parts ) when you link to them. Some of us are busy here. I can't and don't want to read 157 lines of text about how to make AAA games cheap. I like to know about it, I'm interested. I just don't want to read the whole fucking article. Thanks.

"Cut down on redundant staff" was what I got from a skim

Any hobo can make a AAA game. It's all about AAAA games now.

And sadly, unlike women and grades, games do not get better the closer to "A*∞" they are.

 

[sea]

This is a great post but I think it does exaggerate things a little bit for dramatic effect.

High-quality art assets take an incredibly long time to create. The main character in a game might go through five or ten stages of revision even once concept art is hammered out. Minor characters can take weeks or months of work. Animation, motion capture, cutscenes, etc. all take months, even fairly short ones, and require huge amounts of planning and coordination to pull off (because you're probably working with other studios to do a lot of it). Recording dialogue can take months, and even a random shooter has a lot - this can only be done once you've

Now to be fair, I totally agree that modern game development is bloated - there's a hell of a lot of bureaucracy, everything must be approved and finalized and checked and double-checked by the publisher, marketing, lawyers, the producer etc. before it enters full development, and QA stuff must also be very strictly documented - you don't fix a bug without filing some sort of report about it so that others won't go ahead and repeat or break your work. It gets very excessive, sometimes to the point of absurdity.

However, think about a game like Call of Duty, or Assassin's Creed, or whatever. Those have thousands upon thousands of detailed art assets which take years to produce and tweak to get looking right, even with a large team. To make game levels as shiny and interactive as they are, they go through multiple stages of iteration, additional detail passes, etc. Placing those hundreds of interactive objects by hand takes time, after all. What about all the complex scripting that takes place? A single 15 minute sequence in Call of Duty has more raw work effort poured into it than most RPGs have in 6+ hours of gameplay.

It's also worth noting that this is very, very variable. Many developers have smaller teams of 50 to 80, and they still produce "AAA" games. Their employees tend to have more autonomy and not every stage of the project is explicitly micro-managed. Many developers create high-quality titles even with very small teams, and developers do all sorts of different tasks. I recently interviewed for a level design position (didn't get it ) and was told that I'd basically be working on everything from the core level design, to doing scripting and triggering of events, to writing dialogue, to contributing to other aspects of the game design. That might not be true at all at another studio depending on how it's structured. So yes, EA might have some guy working on modeling noses of Madden players for 2 years straight, but that simply can't be said of many or even most other studios.

 

[Cowboy Moment]

I'm kind of torn on this issue. On one hand, the headcounts on development teams for AAA games are ridiculously excessive. Windows 7, which is as large a software project as one can currently imagine, had roughly a thousand people working on it. So, why the fuck does Assassin's Creed 2 need 450? What exactly is it about Epic Mickey 2 that requires the attention of nearly 200 people? There's no way this is appropriate, given the quality and breadth of the games.

On the other hand, I think comparing game development to movie development can be deceptive. Sure, you don't need five writers and three managers for a CoD clone, but it's naive to think one person can maintain creative control over a game the same way the director can author a movie. Perhaps the problem comes from the fact that games use software development processes, which aren't meant to encourage authorship. It's like if someone tried to produce a movie the same way they design a house for an unspecified client - the result probably wouldn't be very good, because the methodology doesn't fit the medium.

One thing I'd also like to point out, is that a huge percentage of people involved in any software development project - and this is even more true for games - will be the QA people. Polish (and I don't mean potato juice) on a complex piece of software costs A LOT of money and time. Even in the smallest of indie projects - AoD - Vince and co have spent most of the past year tweaking and polishing stuff. Incidentally, this is the one useful service publishers provide - they have extensive and experienced QA departments.

A script for a two hour movie has 120 pages. At the speed of three pages a day I can write it in two months and rewrite it completely 6 times within a year.

Here's what I always wondered. Why would you ever want more than one writer for a game? So things can seem choppy and inconsistent and nonsensical? Hey we really need a fat housewife who writers gay porn on the team, that will make the next game better.

It's also about the sheer size of game scripts - most easily seen in RPGs. 120 pages of script with typical formatting is about 30000 words. The script for PS:T clocks 800000 words. There's no way a single writer can produce and revise all that.

 

[sea]

One thing I'd also like to point out, is that a huge percentage of people involved in any software development project - and this is even more true for games - will be the QA people. Polish (and I don't mean potato juice) on a complex piece of software costs A LOT of money and time. Even in the smallest of indie projects - AoD - Vince and co have spent most of the past year tweaking and polishing stuff. Incidentally, this is the one useful service publishers provide - they have extensive and experienced QA departments.

Bingo. Any developer with some experience can churn out a game in a year, but making a game that is free from major issues, outstanding bugs and glitches, and is actually fun to play is a very different question - heck, even small things, like getting the heads of characters to move properly in a scripted sequence, or animations to transition smoothly, require tons and tons of effort that players only notice if something goes wrong. Call of Duty without the polish and tweaking plays like... well, like Fallout 3's shooting.

Design and engineering only goes so far on paper and most of development time comes down to making everything flawless, not just building levels and dropping a few AI characters in there. There's a reason why modern games play so smoothly, have such good UIs and controls, cameras that don't get stuck on walls etc. - it's because they've been tweaked to oblivion just as long as the game was in development. It's making a game where you *don't* really notice any major problems that takes so much time and effort, not just making the game in the first place.

 

[SCO]

That's true of all software. Maybe slightly worse in games since they are meant to simulate worlds and hide stuff from you for a while.

 

[Roguey]

His dismissive comments about game writing are concerning considering the open world with stats-thing they're making. Why settle for Bethesda-level writing in a prettier looking world?

Also I wouldn't be surprised if those Harry Potter movies had script doctors, they're never actually credited.

 

[MetalCraze]

Actually here's a related short quote from the recent Brian Reynolds interview

http://www.gamesindustry.biz/articles/2012-03-23-game-industry-legends-brian-reynolds

That's why the triple-A industry frankly invented expansion packs and downloadable content, because that's what gives half the team something to do while the rest of the team is making the next game.

Really they have too many fucking asshats getting money for doing nothing so that Bobby Kotick can come out and say we spent 60 mln bucks bros.
And of course when you have former Pepsi-Cola manager as a head of EA expenditures will be 10 times bigger than they really should be to achieve the same.

High-quality art assets take an incredibly long time to create.

Console assets are forever stuck in 2004/2005. Graphics doesn't change but games became shorter. At the same time teams grew.

They produce 2 times less assets of the same quality than 7 years ago but have two times bigger teams than 7 years ago.

Also doesn't seem to bother Bohemia or Larian which are p. small. CDP too.

 

[sea]

They produce 2 times less assets of the same quality than 7 years ago but have two times bigger teams than 7 years ago.

I dunno. Call of Duty (for example) has hundreds of unique character models and pieces, hundreds of gun models (first person and third person, mods on weapons, etc.), tens of thousands of models for environment art, etc. Every gun, piece of clothing, hairstyle = new high-detail model and texture that takes longer to create than it used to, because you can't get away with sticking a few boxes together anymore. Same goes for say, destructible environments, dismemberment, and so on. Modern console assets may not be much higher quality in terms of texture resolution and polygon count, but asset streaming systems mean they have far more variety than they used to, which requires far more base artwork - can't get away with tiling 2 textures over and over anymore.

 

[Damned Registrations]

I'm inclined to agree with the article. Aside from the facepalmingly obvious lack of need for dozens of writers to write 300 pages of script over 3 years of development, things like making assets are also easier in a lot of ways. You can't tell me the difficulty of making textures scales linearly with their complexity or size- we have insanely good tools these days, photoshop version umpteen billion and various 3D tools, scanners, google, various online resources (not to mention resources of already owned assets used on previous games). Some bloat is reasonable, but nowhere close to the extent that has occurred. I still recall playing Goldeneye on the N64 and picking one thing out of the credits that struck me as totally fucking insane- elevator music. Someone on the team specifically did the fucking ELEVATOR MUSIC. Yeah, that creative endeavour must have warranted thousands if not tens of thousands of dollars. I'd imagine if the game was made today they'd have different people doing music for each individual elevator.

 

[Dim]

On one hand, the headcounts on development teams for AAA games are ridiculously excessive. Windows 7, which is as large a software project as one can currently imagine, had roughly a thousand people working on it. So, why the fuck does Assassin's Creed 2 need 450? What exactly is it about Epic Mickey 2 that requires the attention of nearly 200 people? There's no way this is appropriate, given the quality and breadth of the games.

This is 2/3rds of the article . I agree. More proof is what people make for free;
or for oblivion or just look at moddb.

 

[sgc_meltdown]

District 9 - 30 Million
Transformers 2 - 200 Million

Witcher 2 - 8-14 Million
Skyrim - 100 Million (estimated)
Kingdoms of Amalur + Copernicus MMO - 50-100 Million

Apparently polishing everything to a high degree of shine and no rough edges anywhere costs tens of millions of dollars.

A single 15 minute sequence in Call of Duty has more raw work effort poured into it than most RPGs have in 6+ hours of gameplay.

yes but raw work here means production assets and making sure the interactive cutscene train isn't derailed while the player follows the onscreen instructions with his thumbsticks, i.e. out of the player's hands one track bullshit like the truman show of games

you might as well say that michael bay films have more raw work effort poured into it than most dialog driven psychological crime dramas have in an entire season

A 100-hooker papal orgy cum buffet with slaughtered endangered animals probably costs a lot more than an excellent dinner+standup comedy tickets+professional escort, but is the expense justifying itself just as much on the former because 'services and goods were rendered' on both ends, so to speak?

Likewise, I really do think nextgen game budgets are more for general perception of luxury than focused goal-related utility or need.

There's a reason why modern games play so smoothly, have such good UIs and controls, cameras that don't get stuck on walls etc. - it's because they've been tweaked to oblivion just as long as the game was in development.

tweaked to oblivion you say
yeah for this example tell me how much of this tweaking went to the famously broken ps3 port or pc ui and hilarious memetastic bugs

and what about the famous Crysis 2 barricade with a zillion tesselated polygons

come on bro I don't think good polish or recognising that a camera needs work is a reason that studios spend as much money these days as you like to think, it's all about the makework and bloated asset pool

in closing I feel the point of the article is to
1) point out the diminishing returns associated with AAA budgets
2) how good knowledge of what needs to be done to make a complete game without the chaff can reduce even the requirements of something that equals the 'enjoyment' of AAA and cost far less

 

[Wirdschowerdn]

Creating the actual assets isn't that much of a time effort, but once you apply it with physics, animations, shaders and other shit so that it behaves properly when it interacts with the rest of the game, problems arise quickly.

Character driven games particuarly are very expensive to make (duh).

Remember, the original BioShock was created by ~50 people (minus PC port). It sold something like 2.5 million units within a few months. Take a look at the game. Not much character variation, animations were crude and basic, no direct interaction a'la Mass Effect, no facial naimations needed, no complex interaction with the environment (except little sister/big daddy)...and whoopy, you don't need to hire another 100 animators, cutscene designers and other "experts".

If I'd ever end up making a friggin' game, it would propably be a dungeon crawler a'la Unltima Underworld or Dark Souls. You can keep the team relatively small yet the shit has potential to sell at least a decent amount of copies.

 

[circ]

District 9 - 30 Million
Transformers 2 - 200 Million

Not really the best of comparisons as D9 is up there with shitty movies along with Transformers. Never saw Transformers 2 as I had enough of derp with 1. Also D9 had a tiny budget compared to Transformers and its massive ad campaign.

 

[Wirdschowerdn]

I like your avatar, circ. Very chic.

 

[attackfighter]

tweaked to oblivion you say
yeah for this example tell me how much of this tweaking went to the famously broken ps3 port or pc ui and hilarious memetastic bugs

and what about the famous Crysis 2 barricade with a zillion tesselated polygons

come on bro I don't think good polish or recognising that a camera needs work is a reason that studios spend as much money these days as you like to think, it's all about the makework and bloated asset pool

in closing I feel the point of the article is to
1) point out the diminishing returns associated with AAA budgets
2) how good knowledge of what needs to be done to make a complete game without the chaff can reduce even the requirements of something that equals the 'enjoyment' of AAA and cost far less

Skyrim had tons of content to justify its development cost.

 

[Awor Szurkrarz]

Witcher 2 - 8-14 Million
Skyrim - 100 Million (estimated)
Kingdoms of Amalur + Copernicus MMO - 50-100 Million

Apparently polishing everything to a high degree of shine and no rough edges anywhere costs tens of millions of dollars.

Take into account that the Witcher 2 used cheap labour from the labour camp of European Union.
It's very easy to have a small budget when you have workers doing 200% of the norm for 1/4 of the money.

 

[Cowboy Moment]

Witcher 2 was also made by a team of ~40 people, so it's a good example in favor of Vavra's point. I imagine not having retard executives without any understanding of games fucking with their development process helped a bunch too.

Btw. that 100 million for Skyrim seems too high. Is that supposed to include marketing costs?

 

[evdk]

Witcher 2 - 8-14 Million
Skyrim - 100 Million (estimated)
Kingdoms of Amalur + Copernicus MMO - 50-100 Million

Apparently polishing everything to a high degree of shine and no rough edges anywhere costs tens of millions of dollars.

Take into account that the Witcher 2 used cheap labour from the labour camp of European Union.
It's very easy to have a small budget when you have workers doing 200% of the norm for 1/4 of the money.

From the stories I have heard about crunch time at big development studios I can't quite imagine how would you do 200% of that without dying. No disputing the working for potatoes part though.