Developer of 'Steam Marines' talks sales, Linux represented 2% over the lifetime of it

Quoting: m2mg2Thanks for the additional detail and for supporting Linux! I honestly probably haven't and won't buy any of your games but I still appreciate your work. It's unfortunate to hear so many bugs on steam marines 2 using Unity, does this appear to be more work than it was with 1/no Unity? What version of Unity? Is it actually Unity or plugins/middleware that are the problem? Do you plan on releasing Mac/Linux day 1 with Windows? Comparing 3 years of Windows sales to 1 year of Linux sales doesn't seem very helpful. Would be interesting to see if the numbers are the same with simultaneous releases.

Np. Don't worry about buying my games, I'm doing fine! Also there appears to be at least one person here who'll be mad at me regardless of what I say...

I'm only halfway through development but for now it seems unlikely that a Linux build will be as difficult, as a percentage of effort and time, compared to the previous game. That said a benefit of Torque2D was that I had access to the engine source code; I have not paid for access to Unity's so in theory something could come along and hose me.

A large source of problems in SM2 as opposed to SM1 is shaders. Stuff that just breaks completely depending on hardware and drivers and what render path you're in. I have a game breaking bug with my fog of war regarding Linux machines on Intel HD graphics. I actually use the Windows Unity Editor for all Win/Mac/Lin builds which cuts down on the Linux editor-specific bugs. Unity has an active issue tracker. I'm currently on 5.4.1 - I'd prefer not to have to freeze my engine version if possible.

You're right that it's not completely fair to compare differing time frames of Linux to non-Linux sales, but that's also implicit in the cost of having a Linux build in that particular case. If it was costless to have a Linux build upon initial release that would likely have improved Linux sales, but it wasn't because of the time factor.

It's unlikely I'll release Steam Marines 2 with WML builds on day one just because of the volume of issues I'd have to work through. The current plan is to launch the Windows build first, hammer out any outstanding issues, then bring Mac and Linux builds to parity.

Quoting: WorthlessBumsAlso there appears to be at least one person here who'll be mad at me regardless of what I say...

I m not mad at you,you're free to whatever you want to say.You're free to support Linux,Mac or not.

I m mad about some naive Linux users.They're just utterly optimistic,which is insanely hilarious.

Quoting: MaCroX95@WorthlessBums

I see that you've taken the correct approach this time by developing a new game in Unity :) There you can literally just push compile and the game has a great chance of already working on all platforms, just avoid using platform specific middlewares and you should have no time or cost loss there :)

Secondly, the testing. Trust me when I say that www.gamingonlinux.com is one of the best and nicest communities in general, sure there are exceptions, but make this community your friend, all in all it is great and trust me that a lot of users would love to do majority of the testing for you as well and report the bugs for the linux platform so early access or beta could be a great Idea regarding Linux. Even some Linux youtubers would probably love to do a review of your games if you let them know about them and promoting it!

Not really sure about how Mac stands, a lot of devs say that it is much more profitable to support Mac eventhough I honestly think that this is just because of the commercialism, Majority of Apple's hardware cannot even run the newest AAA games and those that can cost so much money that only small portion of people can afford it... so percentage of marketshare for Mac can be really misleading as long as AAA gaming is concerned.

Best of luck on your new game and let us know when you finish it!

Yes, I must also admit that C# is much nicer than Torquescript and C++ :D

Like I mentioned in my last reply to someone else, there are platform specific issues to work through like shaders and display driver problems and so forth, so it's not completely frictionless. I'm new to shaders, so even tiny roadbumps can be confusing!

GamingOnLinux has been suggested to me before when I was testing SM1 for Linux. That said I didn't encounter too many technical issues so it just didn't happen. It is more likely I'll take advantage of this in SM2.

SM2 performance on Mac is... not great. I recently bought a new 27" iMac (late 2015 model?) and it's completely fine for coding or watching Netflix or something, but it does not run games well. It's been a struggle optimizing SM2 for Mac.

The general expection seems to be Apple will give its hardware a kick if only for VR because as it stands it's unreasonably bad.

Quoting: WorthlessBumsAlso there appears to be at least one person here who'll be mad at me regardless of what I say... ...
I have a game breaking bug with my fog of war regarding Linux machines on Intel HD graphics. ...

You're right that it's not completely fair to compare differing time frames of Linux to non-Linux sales...

It's unlikely I'll release Steam Marines 2 with WML builds on day one just because of the volume of issues I'd have to work through. The current plan is to launch the Windows build first, hammer out any outstanding issues, then bring Mac and Linux builds to parity.

Hey, first of all thanks for prompt explanations, it's always scary in the dark for the passionate Linux lovers. Second - ignore the guy, he must be drunk. Third - I dare you to make an early access technical demo build for Linux, might be sold through steam as pre-purchase for the real game if you want some money flowing . Don't worry about bugs, let them be, people know what they sign for and most of all might even help you somehow. You know that here and phoronix is a nest of devs, (including Intel, Nvidia and AMD devs) and they might just become interested in some bugs. If it is not against steam guidelines, that is.

@WorthlessBums, great to have you here :)

I would like to comment on a point you raised here:

Quoting: WorthlessBumsEstimating is hard. I'd like to address the entire "2% is really good because we're 1% of the user base" argument. Not every game on Steam has Linux support. You'd expect the average number of Linux sales to be higher than the average percentage of Linux users, per game that does support Linux. If Linux was 10% of Steam's userbase but there was only 1 Steam game with Linux support you would not expect that game's userbase to be only 10% Linux users; the Linux users are simply flowing into fewer games than Windows users are. That said the Mac percentage numbers are low as well - maybe I'm missing something here.

I don't agree with the argument that Linux having less games, means each game will have more sales. We are still normal people with limited incomes, so we won't buy literally everything, we still have to pick and choose what games we are interested in to pick up. Just having less options, doesn't mean we are likely to buy more, it just means we have less to choose from (and in certain genres like this, we have tons), I don't see it any other way.

I do very much appreciate the fact that you will continue to support Linux, don't let the minority of bad attitudes sway you to walk away (you get them in all communities, just the Linux community is smaller so bad attitudes can be a bit louder).

WorthlessBums: I presume some people are a little on edge here because this kind of bad news just doesn't emit the signals to other developers that we'd prefer. I see your numbers are actually lower than my estimates (I'd expect 20k-40k for a port, due to time used, things that have to be learned, etc) and the fact most of the post release support is just staying compatible with the distributions should be music to these people's ears.

At the same time, as I stated before, your game doesn't look like something that would be interesting after a year of playing (I'm saying this from a distance, I haven't really looked at it as I skip most 2D games). If people downloaded your game or bought it on windows, it's going to be unlikely to see a return on the Linux version if you release it a year after. This has been one of the more common gripes with the AAA industry as it takes them over a year to port their PC game to Mac, while it takes the development team that same amount of time to release a new version of that game that then also needs a year to be ported to Mac (and this isn't even talking about Linux).

So if you're doing Steam Marines 2, I'd prefer to see a day 1 launch so we can see a more accurate comparison for the above reason.

Quoting: inlinuxdude1) I'd like to thank the dev for supporting Linux and also for his thoughtful post here on GoL..

I agree. I really appreciate the interesting development comments. I don't play roguelikes, but I always admire people who put the work in to multiplatform games.

Quoting: inlinuxdude2) Is there a way to ignore people like Leopard in the comment sections. His comments really make me sick.. They are singularly unhelpful in attaining our end goals of game parity with other platforms, as well as unappreciative and ungrateful.. :(

I support this. Sadly, a user "ignore" feature wasn't something I worried about until today.

Hey WorthlessBums

To complete what MaCroX95 said earlier about testing, you can also reach our subreddit: https://www.reddit.com/r/linux_gaming

Sometimes studios ask there for linux testers.

Here are last threads that come into my mind:
- Our studio is currently developing a Linux build for our game, Tuebor - Anything we should know specific to Linux gamers?
- We are looking for beta testers for our game Redie.
- Help testing my Linux build
- Need Linux beta testers for our upcoming turn-based strategy game
- My new game is on Steam! Free keys for testers!

Thanks for the response WorthlessBums.
I feel that what doesn't help the sales numbers partially are situations like mine where I don't have a big budget and have a excess of games, so I started to prioritize games from developers that were bought previously, high review scores (or recommendations) and the AAA releases rather than a lot of indies, which translates to fewer games acquired per year.

Hi WorthlessBums,

Thanks for your informative post and Linux support.

Quoting: WorthlessBumsDeveloper here ;)
Estimating is hard. I'd like to address the entire "2% is really good because we're 1% of the user base" argument. Not every game on Steam has Linux support. You'd expect the average number of Linux sales to be higher than the average percentage of Linux users, per game that does support Linux. If Linux was 10% of Steam's userbase but there was only 1 Steam game with Linux support you would not expect that game's userbase to be only 10% Linux users; the Linux users are simply flowing into fewer games than Windows users are. That said the Mac percentage numbers are low as well - maybe I'm missing something here.

While there are fewer games for Linux, there still are over 2000 now to compete for sales. And I am sure it goes for Linux too that a few bestsellers count for the bulk.
There might be a formula to get a "expected" sale, but I am sure its not a simple one.
A gamer looking for a new game will not put an equal chance on all available games not owned yet and do a roll to decide.
We decide after previous experiences, taste, whats getting positive mentions from friends and in the medias we follow etc.

And for what its worth you getting mentioned here on GOL and posting in the thread got my attention and I look into the game now. No promises that I buy it though:-)

Quoting: EdmeneThanks for the response WorthlessBums.
I feel that what doesn't help the sales numbers partially are situations like mine where I don't have a big budget and have a excess of games, so I started to prioritize games from developers that were bought previously, high review scores (or recommendations) and the AAA releases rather than a lot of indies, which translates to fewer games acquired per year.

Same here, I buy mostly bigger games. I do buy a few indie games but I'm pretty picky about them and my taste is a little particular. I'm quite fond of Slain and Mother Russia Bleeds.

Quoting: liamdaweI don't agree with the argument that Linux having less games, means each game will have more sales. We are still normal people with limited incomes, so we won't buy literally everything, we still have to pick and choose what games we are interested in to pick up. Just having less options, doesn't mean we are likely to buy more, it just means we have less to choose from (and in certain genres like this, we have tons), I don't see it any other way.

Honestly the empirical evidence here seems to indicate your conclusion, if not necessarily your exact reasoning ;)

There could definitely be reasons to expect a lower Linux percentage than user percentage. It could well be the lack of options or something else. If we assume that the smaller potential library is a net negative (not unreasonable - a larger library is a significant reason why Steam is the big dog, right?) this doesn't really bode well for developers making games for Linux, or Linux gamers for that matter.

It really does beg the question if there are that many Linux gamers ready to buy these non-ported games, are developers leaving a lot of money on the table by not porting? As someone else mentioned my SM1 porting costs were actually extremely low, largely because of how I run my company. That there are Mac companies like Aspyr suggests that at least large non-Mac games publishers believe porting is both costly and risky.

I don't have the answer.

As a Linux User I recall having absolutely no interest in this kind of game.

To be fair - the developer feels disappointed about their expectation of sales for Linux & Mac platforms.

This thing sells for $20? :\ StaeamSpy says that there were 111,707 ± 8,641 sales.

Is it true those versions came out 1 year after the Windows version?

So the combined 5% is equal to $55,853.50 of extra income in 1 year or $4,654 a month.

If I was a game developer I wouldn't minimize 55 thousand dollars just because my game netted $2,233,022.93 (2.3 mil)

Lets not forget that porting to Mac & Linux is the first step in deploying to XBONE, PS4, iPad/iPhone, Android, etc... when the game has those targets in mind from the start.

Let's not also forget that 1% of the United States controls the vast majority of the wealth in the country.

Linux is similar - our 2% is comprised up of very intellectual people, people who make youtube (Bryan Lunduke) videos, programmers, bloggers (such as Liam), IT's, System Admins, and people who are highly regarded in recommending hardware and software to their piers and families.

I single-handedly am responsible for at least 30-50 people switching to Nvidia instead of AMD/ATI in the last decade. As a Professional IT I would make recommendations on hardware nearly on a weekly basis.

Linux and Mac users are a cult following and highly evangelical as activists supporting or opposing specific games, apps, etc...



View video on youtube.com

An important lesson that judging with your eyes or purely by a mathematical number is deceptive.

Here's a video showing off non-eclidean level design, have an open mind.
View video on youtube.com
View video on youtube.com


Last edited by ElectricPrism on 8 November 2016 at 9:55 pm UTC

Quoting: reaVerWorthlessBums: I presume some people are a little on edge here because this kind of bad news just doesn't emit the signals to other developers that we'd prefer. I see your numbers are actually lower than my estimates (I'd expect 20k-40k for a port, due to time used, things that have to be learned, etc) and the fact most of the post release support is just staying compatible with the distributions should be music to these people's ears.

At the same time, as I stated before, your game doesn't look like something that would be interesting after a year of playing (I'm saying this from a distance, I haven't really looked at it as I skip most 2D games). If people downloaded your game or bought it on windows, it's going to be unlikely to see a return on the Linux version if you release it a year after. This has been one of the more common gripes with the AAA industry as it takes them over a year to port their PC game to Mac, while it takes the development team that same amount of time to release a new version of that game that then also needs a year to be ported to Mac (and this isn't even talking about Linux).

So if you're doing Steam Marines 2, I'd prefer to see a day 1 launch so we can see a more accurate comparison for the above reason.

Unfortunately my results are more or less in line with what a lot of other (indie) game developers have intimated to me privately and publicly. I mean we would also be happier if our Linux ports paid off more.

You're referencing behavior I haven't really seen evidence of: that people would buy a game for Windows and a year or so later buy the same game again but for Linux. I've only seen this behavior when there is a full re-release like a special edition or GOTY or something of that nature, something with a substantial update and not merely a port.

I admit I am curious if Linux builds would do better if launched on day one, but my feeling is that it's unlikely given that Windows and Mac builds of SM1 launched simultaneously and Mac sales are still underrepresented. I'll see what I can do.

Quoting: WorthlessBumsEstimating is hard. I'd like to address the entire "2% is really good because we're 1% of the user base" argument. Not every game on Steam has Linux support. You'd expect the average number of Linux sales to be higher than the average percentage of Linux users, per game that does support Linux. If Linux was 10% of Steam's userbase but there was only 1 Steam game with Linux support you would not expect that game's userbase to be only 10% Linux users; the Linux users are simply flowing into fewer games than Windows users are. That said the Mac percentage numbers are low as well - maybe I'm missing something here.

Maybe :-)

1) If it is true that the Linux version arrived one year late to the party, then you probably sould not expect high sales due to marketing campaign long gone.

2) Did you advertize the game on the thematical resources? Or those who bought the game had found it themselves? We have more than 5 games available under Linux now and it is getting hard to know them all. :P

3) Humble Indie Bundles? Other stores besides DRM-ed Steam? You know, DRM-hatred is especially strong under Linux.

Quoting: WorthlessBumsIs 2% good or bad? If it's bad then there's presumably things I could do to get more Linux users.

Judging by Humble Indie Bundle sales, it is rather weak result, especially for a small indie project.

The recomendations are ordinary:
1) aim for day-1 release
2) if 1) is not possible, jump onto some "Mac and Linus debut" on Humble Sale or something similar
3) differentiate selling stores. I believe it is rather easy to get lost on Steam in that big pool, especially with (as was reported earlier) the "Linux only" filter permanently broken.

Quoting: WorthlessBums...I'm making the game in Unity instead of Torque2D. ... and I have Linux-specific bugs a mile long.

NO! It is not you who have Linux-specific bugs! It is Unity3D game engine. And if you are using commertial version then Unity3D-devs are obliged to provide you working engine. Go, hit them with something big and heavy! All of the Linux community will be grateful!

Quoting: WorthlessBumsIt's unlikely I'll release Steam Marines 2 with WML builds on day one just because of the volume of issues I'd have to work through. The current plan is to launch the Windows build first, hammer out any outstanding issues, then bring Mac and Linux builds to parity.

Than you should expect under-par sales. You know, all my friends are under Windows and I kinda like to play the game when they play it to share my opinions, for multiplayer etc. ... and not one year later when they already have forgotten about it.


Last edited by Alm888 on 8 November 2016 at 10:15 pm UTC

Quoting: ElectricPrismThis thing sells for $20? :\ StaeamSpy says that there were 111,707 ± 8,641 sales.

Is it true those versions came out 1 year after the Windows version?

So the combined 5% is equal to $55,853.50 of extra income in 1 year or $4,654 a month.

If I was a game developer I wouldn't minimize 55 thousand dollars just because my game netted $2,233,022.93 (2.3 mil)

You forget about sales and the cut that Steam and other shops like Humble take. And I believe the prices around the world differ a lot from US and EU prices too. So the numbers he got a probably a lot lower.

I wouldn't just look at the gross sale numbers in this regards. First of all it's always hard to figure who buys for what platform or if someone wouldn't have bought a game if it was/was not available for Linux etc... So real benefit in terms of money is hard to get right imho.

But there's also another point in supporting OSX/Linux: it shows interest in the users demands, often gives positive public reception, sometimes even (indirect) advertisment like in this case.

As Unity and its "just push Linux export" was already mentioned here... It's obviously not that simple yet. But I'd argue it's very close to that. Most of our issues have always been shader/graphics related, as WorthlessBums also said. But unfortunately that's not a minor issue as debugging that stuff is very annoying, even more so with theamount of HW/driver/distribution combinations.

Of course there's other stuff too, like file I/O (can be handled quite easily if you do care about a few things though), locales (the Linux player of Unity is the only one which cares about the system locale, resulting in issues with float/datetime parsing so you have to explicitly use culture invariant parsing), or RNGs behaving different. Nothing taking too much time in my experience.

Native code can also become an issue if you use stuff that's not available for those platforms bjt you can typically avoid that. Since like half a year now even the Steam API can be fully used, including Linux x64 builds (was crashing on some parts before).

But after having all that sorted out ... you still won't get only happy users. There's seemingly random crashes for some (luckily the minority) in the Unity players without any useful feedback for a dev in the logs. That's where working on that platform really gets annoying, even if you really want to get it to work (like myself, been working on the Linux support even before I was part of the company ;) ).



All in all I'd say with an engine like Unity3D supporting Linux/OSX is definitely worth it, as the required work to get it working for the majority of those users is *relatively* low and you get happy (/happier) customers in return.

Quoting: Alm888NO! It is not you who have Linux-specific bugs! It is Unity3D game engine. And if you are using commertial version then Unity3D-devs are obliged to provide you working engine. Go, hit them with something big and heavy! All of the Linux community will be grateful!

lol`ed
Hate to have to say that but paying for a product nowadays is in no way helping you getting support. Made that experience with both Valve and UnityTechnologies. Getting replies at all is hard enough, even more getting actual help. (Tak's a really big exception here but he's "just" responsible for some Linux stuff, not everything ;) )
Not even being able to talk to anyone makes hitting them with something big harder as well.

I think I got this in a bundle, so I probably didn't even count as a linux sale :\

He sounds a little sore that "linux users claimed a bunch of people would buy it if I ported it" -- 2,000 is a good chunk of people. I doubt if that game had a "linux version?" forum thread with 2,000 +1's in it.

QuoteHate to have to say that but paying for a product nowadays is in no way helping you getting support.

Yeah. Never had a good experience with IT/software support. They usually told me to try different computer. Like I am buying a new computer $1000 for $20 game that I owe for half a year and it broke after update.