Game Profitability Under Threat

The BBC has up an article looking at the dwindling opportunities for profit on games in the coming years. Soaring prices for game development, the increasingly-entrenched segmentation of the marketplace, and overwhelming emphasis on sequels means that it's looking increasingly dire for game development houses. While the success of the DS means that there's a wide market for games on that platform (witness Square/Enix's movement of the Dragon Quest franchise), the phasing out of the PS2 means that for the moment there is no 'leading platform' for game creation. The article talks about how the various game companies are responding to this challenge, as in Microsoft's reliance on exclusive deals and Sony's absorption of development houses into their infrastructure.

Comments

the market will sort itself out

[Astarica]

A game is not entitled to make money due to some intrinsic 'goodness' value. If I spent a billion dollar to make the greatest game ever, I would expect to lose money on this because I don't think you can physically sell enough copies to make up the development cost. This means even the greatest game ever is not worth spending a billion dollar on it. If you make a really cool game that no one bought, maybe it's not as cool as you thought it was.

Change in Business Model

[p0tat03]

This necessitates a change in the way games are made. One of the fundamental problems is that we're building games in bigger, shinier forms, without streamlining our method of production. As our graphical capabilities increase, we will be tempted to include more and more content into our games. That 900-polygon character that took an artist 1 week to create now takes 2 artists 3 weeks, what with technologies like parallax mapping.

I have believed, and still believe, that procedural content is the answer. There is a limit to how much manpower a development team can consume whilst remaining profitable, and IMHO we're already at that line. We need to start letting the machines figure out our content. This does not necessarily mean complete and full generation of assets by algorithms, but rather that our tools need to be streamlined as such. Software like ZBrush have drastically reduced (for the skilled user anyway) the amount of time required to build high-poly models. We need more tools like this for textures, for all other aspects of game development. We need to let go of the manual shift stick and build more powerful tools that will take more off our coders' and artists' hands.

This also means the segmentation between game developer and technology developer. For years we've seen some companies stick stubbornly to building their own engines, a costly affair. It should be clear to developers by now that, if you are in any way serious about graphical horsepower, you need to license an engine. Building your own engine from scratch is no longer feasible if you want to get your game done on-time and on-budget. The industry will, in time, become the playing field of dedicated technology developers who license their engines to developers, much like Valve and Epic are doing now.

The gaming industry is holding onto archaic ideals. It is like the car factory that insists hand-built is better, and refuses to mechanize any aspect of their production. It is now suffering the consequences, and like it or not they will have to change.

Re:Change in Business Model

[p0tat03]

Agreed. I would have suggested a solution had I been aware of one.

That said, it's not about doing the things I'm suggesting and turning an obscene profit - this is about restoring profitability, not about making game development so cheap that everyone is rolling in the dough - that won't ever happen.

IMHO we need technological development in tools. We are still using the same 3D modeling tools, the same texturing tools, and the same map-building tools that we've been using since the release of Half-Life 1 8 years ago. The difference is that what we're doing now is many times more complicated than before.

In 1999 a wall was 2 polygons, and a 256x256 flat texture.

Now, we demand things like geometrically modeled light switches on these walls, power outlets, and other little things that add to immersion. We also need normal/parallax maps on all of this, not to mention specular maps too. We've added so much on top of what we used to do, without once stopping to really, completely rethink the way we interact with our tools.

When I first saw an artist use ZBrush, I was blown away. Here we have something that is smart, it is awesomely predictive, and it reduces the workload of the artist dramatically when it comes to modeling high-detail meshes. We're talking a couple orders of magnitude less time to do the same thing. And all it took was a brand new interface and way to interact with the type of models we're used to seeing.

Let's see the same sort of rethinking for our animation, for our texturing, and for our mapping.

Good news!

[mdielmann]

I've been waiting for this for a long time. Maybe we can get something besides yet another sequel, a movie spin-off, or a blatant rip-off of another game. Perhaps they'll have to settle for plot and gameplay (or at least just gameplay) instead of stunning graphics and no substance. As a proud new owner of a Wii (yay!), I have to say the graphics are good enough, maybe not a match for the XBox 360, but the games are fun. This matters most of all. Of course, new ideas are a risk, while sequels are a known quantity with an established market. I think that's how the marketing goes...