Raph Koster has another post up about game development costs, which continues the conversation we had here back in this post. This time he comes with more data, but I have some strong objections to said data, and again with what it means.

The main data point I disagree with is the cost-per-byte analysis. Hard drives have gotten significantly larger and cheaper over the years, while internet connections have gotten faster. Ten years ago if a game required you to download 20 gigs, and needed 50 gigs on your hard drive, that would have been a limiting factor for many users. Today? Most of us don’t even blink at that; we just hit download on Steam, wait a few minutes, and we are good to go. Sometimes those of us who keep dozens of games might need to do a little drive cleanup, but if you typically uninstall a game once you are ‘done’ with it, you might NEVER have an issue with drive space.

And developers know that, so if they cut a corner and increase the download size by a few gigs, it likely doesn’t matter today. Way back in the day, when game size was capped by the size of a cartridge, you simply could not cut that corner most of the time.

That is just one reason why looking at the size of games today is greatly misleading in my opinion. Another reason is that today, gamers will easily trade having much larger files for very minor benefits. A great example of this are 4k texture packs. Unless you eagle-eye the specific texture, most gamers won’t really notice a difference between a 4k texture and a 2k, or even smaller. But again, since file size is mostly a non-factor now, developers include those 4k textures, which massively increase the total download size of a game, while not having a major (or any) impact on development costs (they already had said 4k texture).

The other point I have issue with is the example of the character model and the amount of work it took back in the day vs today. As models get more complex, they take longer to make, which makes sense. What doesn’t add up is the need for all games to use more and more complex character models (or any models for that matter). A could list dozens of recent, successful games that don’t rely on state-of-the-art graphics. ‘Good enough’ today often times means the graphics of a game are more than acceptable, in part because of how powerful our devices are, and also because not all games need to have a 3D photorealistic look to work. Using standard Unity or Unreal art assets today doesn’t automatically result in your game looking cheap, and with some limited additional work, it can end up looking pretty unique and memorable.

This gets into a large problem I have with the overall conversation too; that we drift from talking about all games and then specifically to AAA games almost interchangeably, which is a mistake. The AAA space is pretty unique, and not just because of costs. AAA studios have been doing very well in terms of profits in recent years, so while yes, their games are getting more expensive to make, its clearly NOT crushing their profits. Additionally, work conditions for developers at AAA studios have, on average, improved since the early days (they still stink compared to other IT work, but they are better), so again while games cost more to make, it seems to be working out for both publishers and developers (and I’d say gamers too, as we have access to more AAA titles today than ever before, to say nothing of our overflowing choices for non-AAA games).

That’s not to say every example is perfect or a success story, but I also don’t agree that this is some industry-wide crisis that is killing game. Quite the opposite; we have more choices of higher-quality games today than we ever did, while developers have more choices (indie, big studio) in where they want to work, and what they want to make. The shift to ‘games as a service’, to circle back to the original post, is also fine, so long as its done right. Predatory MXT isn’t doing it right, nor is crippling your game and it’s design to fit a cash shop. But those are jut growing pains, and we are seeing that both gamers and developers are, slowly, figuring it out.