I just made a post about this in another thread, but it was a tad off topic and the issue is complex enough to deserve its own discussion, I think.

I'll start with a simple and crude analogy: many people have compared "hardcore" games to art house flicks, in that they represent the tastes of the more sophisticated elite of the medium. Assuming we buy this analogy (and I don't, but let's run with this for a second), there's a significant issue on the horizon.

For movies, it's hoi polloi who demand the mega budget action flicks with tons of explosions that costs tons of money to produce. The smaller artsy flicks can justify their lower box office take because they have lower production costs. Conversely, it's the "hardcore" gamers that demand mega budget action games that cost tons of money to produce; a situation where games are the most expensive to produce while simultaneously being amenable only to the connossieur isn't sustainable. That's a problem.

In short, I think we're rapidly approaching a point where "hardcore" gamers are unable to sustain the types of games they want. Because these gamers are precisely the ones who constantly demand better graphics and more complex AI, it is also the group that needs to show the most growth, and it isn't; I hope we can agree that the majority of industry growth seems to be in the more casual arenas. In fact, looking at the total hardware sales from generation 5 and generation 6:

http://vgchartz.com/worldcons.php

We get approximately 144 million consoles for generation 5, and approximately 174 for generation 6. Accounting for population increases, that's somewhere around 10 percent growth in the industry. Certainly not bad... but hardly the explosive growth we saw from generation 4 to 5 (there were 79 million consoles sold in Generation 4, which means Generation 5 represented a 75% increase in overall system sales with population growth accounted for) or generation 2 to 3 (The numbers aren't available here on VGChartz, but I assume we can agree the industry saw enormous growth from the crash generation 2 caused to the NES years). 

However, as we have noted on this site on many occassions, costs to produce games have nearly doubled in the last 5 years, and you can't have 10 percent industry growth and 100% cost increases without some serious economic consequences, and we're already seeing many of those consequences occur.

Okay, lots more to say, but I'll stop there for the moment, simply because I don't want to post more than people can digest.

I think there is a solution here and it is the fact that we are rapidly approaching a point where games do look good enough and AIs do react smart enough that people can't really complain any more. Obviously improvements will still be a big part of selling a new game and this means the industry runs the risk of growing stagnant but I think if we look at the movie industry we can see that they are in a similar situation and have been doing well even with gradual graphical increases over the years.

The next generation of consoles doesn't have to be the huge boost to raw power that this generation was. All it really has to do is stabilize and extend existing architecture while providing a reasonable boost to power. So on the hardware front things aren't so bad. But this translates directly to the software front because if the hardware is designed properly we might have the first situation where a new generation of hardware doesn't require a complete reworking of how to make software.

I think the increased budgets and cost are in part due to the fact that this level of detail is bleeding edge and very new territory which requires some of the best people to come in and pave new roads to do it properly. But as the experience is gained and shared by more programmers and designers I think the cost will come down as new tools and new approaches are developed that streamline the process. This is afterall a continual effort by most design studios to identify project hangups and work to alleviate them for future projects.

I still have to say I agree there is a risk that what you say could come to pass but smart and agile competitors should easily be able to avoid this downfall and that will leave more market from them to soak up as the slower companies die off.

That's what I've been saying for over a year Bod. It especially hurts game companies becuase those exploding cost are related to the systems (PS360) that left to their own devices would actually shrink the market. Despite the "mega" sales on the 360 it simply isn't enough to support multiple $30+ million games when it takes 2 years to reach 13 million gamers. However, most gamers seem to think that video game companies are supposed to lose money to subsidize their entertainment so hopefully they will be disabused of that notion over the coming years.

btw, Nintendo came to this conclusion halfway during the Gamecube era. Many developers told them that graphics were finally good enough it didn't matter anymore, so they took that design into the Wii.

Many smaller/mid companies were munched during last gen like Hudson and Taito. They couldn't make games for the consoles they liked because of high costs.

With the influx of casual gamers and the Wii's success at going for casuals before hardcore gamers first, maybe the industry will take a step back and realize what they're doing before it becomes too late.

Do these costs take into consideration licensing an existing game engine instead of building one from scratch? What about companies reusing or tweaking existing code, tools, engine, etc that were created for previous games they've made? Are we expected to believe companies re-invent the wheel everytime they made a new develop on a console they've already developed a game on?

What about development kits, didn't Sony just cut in half the cost of their's (from ~$20,000 each to ~$10,000) and that's not even taking into consideration the fact that many teams are on their second or third PS3 or 360 game and that cost has already been paid for with their first game.

Lets take Rainbow Six Vegas 2 for example, it's doubtful the team bought all new development kits for the game or threw out all the code from the first (AI, Physics, Art, etc). Or that they forgot everything they learned from the first game so they have to spend time re-learning everything. I'm willing to bet the cost of making Rainbow Six Vegas 2 is less than the first game.

I see $20+ million numbers being thrown around a lot but it seems a lot of games with those kinds of costs are ones in which the engine and everything else for them are being built from scratch or a lot of customization of an existing licensed engine is taking place. In one camp there're games like Heavenly Sword, Metal Gear Solid 4, and Grand Theft Auto 4 in which everything is being built from the ground up. In another there're games like Bioshock, Stranglehold, and Too Human in which an existing engine is heavily modified. But in both cases I'm willing to bet for the ones that get a sequel it will cost less and take less time to make.

Anyways what I'm trying to say it that this issue isn't as black and white as some people are making it out to be.

HD graphics cost more than SD graphics due to texture resolution.

Dev Kit costs are negligible compared to content creation costs.

There's definitely ways to reduce costs, Legend, and I absolutely agree that the best solution is the one you named: middleware.

Yes, we can expect costs to go down during this generation as engines are established and the wheel is firmly invented. I think we can expect the economy to go from terrible this year to mediocre next year; and perhaps the year after that we will be at "pretty good."

But the year after that, 2010? Companies start ramping up for the next generation, and they really ARE forced to reinvent the wheel again. The middle of the PS2's lifespan is approximately when EA saw their R&D costs begin to increase, presumably as they readied themselves for the PS3.

A year or two of "pretty good" profits won't compensate. Something has to change. 

Well the primary cost for game development is time as the more time it takes to make the game the more you pay your developers. So in that regard if we look at a game like Halo, it took a pretty long time to make the game still and it cost a very shiny penny or 3. The same is true of Mario and how long it took to make that game, from what I can tell mario had a smaller team than what Halo 3 had but it still took about 2.5 years as best I can tell. So games that are a sequel don't seem to be immune to increased or increasing costs just because they are a sequel.

I think it depends on how lazy the devs want to be. If they are just adding in new content then sure there probably isn't a whole lot of work to be done and you can probably go ahead and just get the entire project done in less than 6 months. But if you want a good game I think you have to dig down a little bit and actually and improve on the existing engine in a way the player will notice and appreciate.

There is definitely a smarter way to do it with engine re-use as you aren't doing the same things over and over, but I would hate to see devs get stuck in a regurgitation loop also. 

@chadius,

Graphics don't really have a cost per se, other than what you pay your artists. It just takes longer to make quality HD textures than it takes to make quality SD textures.

One point I in Legend11's post that I agree with: we need better information on game production costs. The $25 million game cost may represent 1% of all games while 60% are at the 1 million mark (pulling numbers out of thin air here.) Sadly most companies don't say how much games cost to make. Especially when they don't break even and the bad press could hurt their brand.

@legend11
What chadius said. The costs, except for IP licenses, are going to be more for an HD game than an SD/ED game not matter which way you cut it. True not every HD game costs $20+ million but the cheapest HD game is still going to be in the range of the most expensive SD/ED game.

@sqrl
I agree though mostly from a business standpoint. I think electronics companies are finally coming around to the fact that people are not going to new $3,000 TV's, $400 cell phones, $600 video game systems, $1,000 movie players every 3 years. The only reason they did a few years ago was due to the funny money mortgage game most people were playing though it seems many companies thought it would be permanent. I would be shocked if Sony and MS repeated their core mistake again next gen.

From a technical standpoint though home console games are not near photorealistic graphics or human like AI (although I imagine that one's more of a software issue than a hardware one). If they're smart they've delayed their plans, or perhaps the technology wasn't quite ready, but I recall reading over the last few years about 2K and 4K resolution TV's that would be released in the next few years. Should Sony or MS want the potential should be there technically to release another $600 "supercomputer in a homeconsole", again if they want to. Though I don't follow such topics regularly so the technology may be delayed or scrapped.

@chadius
I wish I could find the article but actually Nintendo was saying the Gamecube was near the limit of what an SD TV could do. I think it was at E3 2001 where Miyamoto gave an interview saying that and adding that Nintendo was looking at different ways to advance gaming next gen (which we now jnow Nintendo was toying around with the first Wiimote protoypes at that point in time).