on Jul 30, 2008
This post is for developers, investors, and publishers of A-list games. (Or is that AAA titles? Whatever. “Games on the shelf at Wal-Mart for $50” is a more accurate descriptor but it’s kind of verbose. Sort of like me.) I know I’ve got at least two or three of you in my audience, hiding amongst the crowd of regular gamers, indie developers, and my fellow curmudgeons. I’m going to have another go at talking you out of your obsessive pursuit of graphics, which at this point makes Gollum’s pursuit of The One Ring look lackadaisical. I realize this is a hopeless task, but it’s no less hopeless and unfulfilling than trying to keep my humble hardware up to date in the face of your skyrocketing system requirements. And since I’m not having a good time I might as well drag you along with me.
If you’re one of those people who is unable to tell the difference between spending money and making fun games then you’re excused. Go back to developing your juvenile plotless cookie-cutter tech demo and don’t trouble yourself with this business.
In the 1980’s, while I was still fantasizing about becoming a programmer and trying to figure out how to kiss girls (or maybe the other way ’round) your average videogame development team was A Guy. Sometimes larger teams might include His Buddy. As we entered the 90’s and the age of Doom we saw team sizes swell to numbers that occasionally made it feasible to play a little doubles tennis. Not that anyone had time for that sort of thing. A few years later and development teams could, if they ever went outside, possibly fill the positions in a baseball diamond. A few years after that and we had teams of 20 or more people and something strange started to happen. Games started getting shorter. When Max Payne blasted his way onto the scene everyone pointed at his short little 10 hour game and giggled. 10 hours? Who is going to pay $40 for a ten hour game?
Well, all of us, eventually. If we’re lucky. But now we can look back fondly at those days. Teams are still getting bigger and games are still getting shorter. It now takes a hundred people to produce content that offers even less gameplay than Max Payne did. Oh yeah: And development time has increased as well, so not only are you paying a lot more people but you’re also paying them for a lot longer before you actually get your game.
This is not a good trend and it should fill reasonable developers with apprehension, because it can’t keep going like this. The number of PC gamers hasn’t really gone up all that much. You’re still aiming for that 4 million units target you were a decade ago. To put it another way, you’re now funding teams five times as large for twice as long to sell shorter games to an audience that is roughly the same size. You can quibble with these numbers if you like, but the trend is there, and it’s visible. If I’ve overstated or understated the problem by some margin it doesn’t really change that fact the PC Gaming is seeing an unsustainable escalation of development costs.
In Wolfenstein 3D, you can fire up the level editor and make a single room in less then a minute. In Doom you might spend ten minutes making a room. Around the turn of the century you might spend a couple of hours on it. Last I heard it now takes several people (usually a mapper, a 3D modeler, a texture artist, and possibly some sort of scripting person) working together for a couple of days to make that room. Yes, the room is very realistic and cool looking, but it takes 72 man-hours to make the damn thing.
Each new graphics generation requires more development hours to exploit it. The jump from 2D spites to moving 3D characters was a pretty compelling one. The leap forward in lighting technology that give us flashlights actually added something to the gameplay. But adding another lighting pass to make sure a doorknob can support multiple specular reflections? Improving the shading and density of foliage by 25%? Are these really worth the development time and the upgrade cost for the user? How many of those customers would you lose if you didn’t take the next step? What if you just stayed where you are right now from a technology standpoint, instead of adding another 50 people or another six months of development time to your game? Sure, you wouldn’t have the latest graphics, but how many sales is that going to cost you? I hear people claim that they can’t attract gamers without the next-gen graphics, but… when was the last time anyone honestly tried?
I don’t know what the hardware breakdown is out there. How many PC owners have which graphics cards? Now, your first answer might be to jump over and have a look at the Valve Hardware Survey. “Oh look! An overwhelming majority of users have NVidia 6000 series or better!” It really gets on my nerves when you do that and if I could I would whap you on the head with this copy of Game Informer I keep handy for when I need a good laugh. That survey is for Steam users, and therefore mostly people who already own Half-Life 2. Which is exactly the problem you keep running into. You keep aiming your game at the same fragment of the potential audience.
Like I said, I don’t know the breakdown, but if it’s not a bell curve I’ll eat my keyboard. I’ve heard that nothing makes a presentation seem professional like a good chart, so with that in mind I give you this:
Maybe that wasn’t as professional as it could have been. At any rate, you keep aiming your games so that users need the female terminator to run the game well, and a T-1000 to run the game poorly. This leaves out all those people in the middle, who are happy to spend a million billion dollars on Peggle and re-skinned Bejeweled clones. I can hear you arguing, “Oh Shamus, those people are all boring and stuffy and wouldn’t want our videogames. Also I need to be hit in the face with that magazine again.” How would you know they don’t want your games? You’ve never made anything that they can run. They haven’t rejected you, because you’ve never given them the chance. By the time Joe Average has hardware that can run your fancy-pants game, it’s long gone from stores and replaced by newer games he can’t run. Keep in mind that most of these peggle-players have no clue how to use torrents. You keep aiming your game at this tiny, pirate-infested group and wondering why sales are so small.
(Also note that a lot of people in the middle are former PC gamers, who did the math and realized that could take the graphics-card money and put it into a console instead of mucking about inside of their computer every eighteen months. Those people aren’t against PC games, they just don’t have any available. With a few exceptions.)
Case in point: Note how World of Warcraft is at least two graphics generations out of date, and yet Blizzard had to buy avalanche insurance just in case their pile of money falls over. Their game looks like Lego Middle Earth and they are kicking your ass. Part of the reason is because their game will run on almost every battered laptop and second-hand computer on the planet. Your customers, on the other hand, need to blow a couple hundred dollars every other year just to run your games poorly.
As far as making a game fun, graphics spectacle is the most foolhardy and inefficient way to spend your money. The fancy visuals are exciting for the first few minutes, but then the user becomes acclimated and desensitized to your razzle-dazzle and they’re left with just the gameplay to entertain them. And gameplay is the one thing you keep cutting to pay for the graphics. That’s like cutting off the top of your head because your high heels make you too tall to fit through the door.
If you would just take two steps back from that accursed bleeding edge and aim for the middle of the bell curve you would discover that:
- You could work with a much smaller team, paying fewer salaries.
- Since you’re not re-writing your tools and changing your art pipeline every time you start a new game, development time will be shorter.
- Your artists will be more productive since you won’t be snatching away the tools they understand for newer, more complex tools. You might find it to be a little easier to make longer / deeper games.
- You’ll have a far larger potential audience.
- You’ll have fewer support / QA issues because you’ll be building on established technology instead of working the gremlins out of the new stuff.
- Better framerates, faster load times, quicker installs.
Even if none of those Peggle-playing goofs embraced your game, you’d still be better off because you spent a lot less money to sell to the same group of gamers you’ve been dealing with for years. You risk so much money on your new pixel shaders, but you’re not willing to risk spending less and see if you can still get the usual suspects to pony up?
I realize I just summed up using a bulleted list, but let me sum up again, just to make absolutely sure I’ve driven my point home: You can spend far less to make a game with more value that can offer a better play experience to a larger audience with less pirates.
And now let me sum up my summary, for the benefit of those in marketing: You can spend less money and make more money.
Me? I’m saving up for an XBox 360. You guys are driving me nuts with this business.
Shamus Young is an old-school OpenGL programmer, author, and composer. He runs this site and if anything is broken you should probably blame him.