The Business Model of Selling Linux Games

I’ve been enjoying the visibility into Linux Game Publishing’s business at the LGP blog. The author is the CEO and Head of Development Michael Simms, and he is very accessible, responding to comments and listening to suggestions. Before his blog was launched, I remember seeing him on IRC regularly in Linux game-related channels.

Months ago, there was a post explaining why Linux games cost as much as they do. If you’re not familiar with Linux ports of popular games, you’d possibly be surprised that they cost as much as a new game…even though the original version of the game on Windows might cost half as much since it has been out for months or years! The basic argument: even if a game has has been out for some time on Windows, it’s new to Linux, and it costs LGP money to port the game in the first place.

At the suggestion of some commenters, he did a one day sale in which prices for most games were lowered. The sale results indicated that while the short term sales increase looked promising, the long term sales stayed flat with less revenue, not more.

But another thing LGP has done at the request of customers is offer downloadable games, including a rental option at a “ridiculously lower price”.

Most recently, there was the question of why LGP can’t simply provide the Linux version for free to people who purchased the Windows version. It’s a legitimate question, especially when companies such as id Software are providing free binaries for all platforms to people who purchased a game. I have a copy of the Linux port of Quake 3 Arena, and I know I can run it on Windows without paying for a second copy. I mention this example a lot, but I know one person who was told that she couldn’t play The Sims and all of the expansions she purchased on her new Mac because EA outsourced the Mac port of the game to another company, and that company was handling sales of the Mac port.

It’s a business model that works for EA, but not necessarily for the customer. And it’s the same with LGP.

As a customer, I prefer being able to buy a game once and play it anywhere I want. Windows, Mac, and Linux. The game is the same. Now, it could be said that I’m arguing that Xbox 360 or PS3 ports of the same game should also be available to me for free after purchasing the same game on my computer, but I’m not. In fact, 360 or PS3 games usually have some exclusive content, making it a different game in some way.

In any case, the argument for the business model is basically the same as the argument for Linux game pricing: you’re paying for convenience, and it costs money to give you that convenience. There is one additional point that Michael Simms made: companies like EA don’t care about Linux, and so the business model I’d prefer is just not going to happen.

Which is too bad. It’s why I find that I don’t purchase many games these days. If I find anything that interests me, I remember that it’s not available for Linux. Instead of supporting those games, I could spend my money on a game that is actually supported where I want to play it. Starcraft 2? Yeah, call me when it runs in Wine. Maybe. I might want to play a different game by then, one that is natively supported.

The good news is that the business model that LGP currently uses seems to be working for them. Micheal Simms said, “The business model we have isn’t ideal, I’ll be the first to admit. But it is the only one that works as the market stands.”

I’m not so sure that it is the “only” model that works today, but I’d like to hear more about other developers and publishers’ experiences. Anyone out there with a different model that works for them?

2 replies on “The Business Model of Selling Linux Games”

My business model is more like you describe – for the games that we develop in-house, you can purchase the game once for one platform, and later if you switch to another platform, just e-mail me and I’ll be happy to give you a copy of the game for your new platform. (This does not work for the affiliate games that we sell on our site, of course, because we have no control over those games).

At one time, Grubby Games did the same thing – I bought Fizzball originally for Windows, then e-mailed Ryan Clark and got a copy for Linux. Of course, they’ve since been bought out by Big Fish, so I don’t know if they would still do this or not.

I think it depends on whether you are developing your own games from scratch or licensing/porting someone else’s game. With the latter, you have several issues. First, there is definitely an investment that you need to recoup (licensing costs, development costs, etc). Second, you have limited rights to your creation – you may have some rights for the port that you create (so you can distribute/sell it), but not for the original game.

It also depends on how the game was developed. Id tends to use either cross-platform graphics libraries or write their own from scratch. That makes the process of porting to a new platform reasonably easy, so for them to offer a Linux binary in addition to a Windows binary is not all that much more work/cost.

A game that makes heavy use of Windows-specific libraries, however, is going to require considerably more work to port to Mac or Linux. That could be made easier by adding an internal abstraction layer, but that’s then more work in development, a performance hit (maybe), and can easily lead to a “lowest common denominator” situation where you end up not able to make full use of your underlying libraries.

While the knee-jerk response is to say “well just use cross-platform libraries”, and I’d love to be able to say that, it isn’t always so easy. Some of the platform specific tools are really really nice, and can save you gobs of time and work to get the result you want… on the platforms that support it. When one platform offers a number of nice time savers like that, *and* controls the majority of the market, it’s not surprising that a lot of companies don’t think about other platforms until it’s too late to do it cheaply.

Just imagine what it would take to port the DirectX-based Half Life 2 to Linux compared to the Custom/OpenGL DooM 3. You’re basically looking at an engine rewrite in the first case, and recompiling and fixing platform-specific bugs in the second. It’s no wonder the second is so much cheaper for the end-user.

Comments are closed.