Steam's top releases of May show why Steam Play is needed for Linux

Quoting: Shmerl

Quoting: BeamboomThere's a lot of talk about market share here. Lots of assumptions and theories. But if there's one thing we DO know, is that the fact that Steam has been on Linux for several years now, and the Linux user base has been, for all practical purposes, completely stationary. Hasn't moved in any significant rate at all.

Total number of Linux gamers affects sales, not market share. From what I've seen, total number of Linux gamers has been growing, year after year. I explained the idea above though. As with classic supply and demand, potential profits depend on both. If there are too many games per certain amount of gamers, profits will go down for developers, even if amount of gamers is growing. So market will balance naturally by lowering the number of produced games. It's not a sign of anything negative.

This of course ends at a certain point. We seem to be at around 25% of games being made/ported to Linux. So if we double our numbers, the equivalent balance point is around 50%, and if we get past 4% the balance point would be at 100% of games. That doesn't mean we'd actually get all the games at 4%, Apple doesn't and they're higher, just that if we did, they'd sell the same as 25% do now. What we'd get is somewhat more games, including more games higher up the expense/marketing slope, and each game selling somewhat more Linux units. But what I'm saying is that above a certain percentage of Linux users, you can no longer get a dilution effect where more games spread the increased sales between them, because we get to "nearly all the games" way before we get to "nearly all the users".

I think this affects bigger markets too. Didn't some developers complain, that the market is overcrowded even on Windows, and it's harder to make profits because of tons and tons of games coming out? In the smaller market this issue is just more pronounced and you can more easily see all the shifts, but it affects all in general.


Last edited by Shmerl on 30 June 2019 at 6:27 pm UTC

Quoting: Guest

Quoting: Linuxwarper

Quoting: GuestUh no, the AAA games we have on Linux were not for charity or good will, we have proof that the work done releasing for and supporting Linux was compensated for.

Fact is if Linux was profitable for triple A and indie AAA (as Ninja Theory calls it), then we would see more than two AAA games for Linux at E3. One of them, the Borderlands DLC isn't even confirmed, I just included it just because Liam implied it would come because of Aspyr. So tell me again if Linux is profitable:
where is Ashes of Singularity. It's been two years. Where is Doom 2016? The port is basically done. Do you really believe those in charge of Doom would not port it to Linux if they could make more money?
Just because Linux has AAA games it does not prove that it's sustainable business as a whole. The sooner you realize that the sooner Linux gaming will be better because of it.

Quoting: Purple Library GuyThe rest of what he said is pretty frightening though. He seems quite definite that Linux gaming is shrinking over time. This over a time period when on a technical level, near as I can figure out Linux has been significantly improving as a gaming platform.
We may need Proton, and Stadia I guess, more than I thought.

It seems quite evident to me. With some exceptions and indies Linux isn't profitable.

Quoting: Purple Library GuyI'd certainly love to see that Windows 7--> Linux migration scenario materialize. But I've seen a couple of other "This MS screw-up should drive people our way" events that never worked out, so I dunno. Time will tell.

I think for that to be effective couple things needs to be in order: anti cheat (EAC and BattleEye), even more driver improvements and gamers on Linux doing free ad campaign for Linux.

Without choosing sides (I haven't dug enough into evidence, data sets, etc, myself yet to make a judgement), quite often there's also the case that supporting GNU/Linux might be profitable, but not quite as profitable as putting that same investment back into the original platform, or to a console, or better marketing.

Of course that concept of opportunity costs assumes that it's not workable to simply hire more people to do an additional project if it's going to be profitable. That's probably a reasonable assumption actually--mythical man month and all that, by the time you get additional hires up to speed it might not be profitable any more. But what it does point to is the role of specialist porters like Feral, people already expert at what you want done who can make it happen profitably without taking up your own developers' time.

Quoting: ShmerlI think this affects bigger markets too. Didn't some developers complain, that the market is overcrowded even on Windows, and it's harder to make profits because of tons and tons of games coming out? In the smaller market this issue is just more pronounced and you can more easily see all the shifts, but it affects all in general.

Seems reasonable; as long as there's a good prospect of making money creating games, people will make more games. Presumably there's an equilibrium point where it's just un/profitable enough for as many people to start doing it as quit doing it.


Last edited by Purple Library Guy on 30 June 2019 at 7:48 pm UTC

Quoting: Purple Library GuySee, this is the assumption I question--that if we just make things better, people will just automatically switch and our share will increase. If we build it they will come.
They probably won't. Oh, there will be a few, there have already been a few and there will be a few more. But that's at the level of anecdotes, not percentages. In the economy we have, and with people being what people are, just having a better technology or even a better product overall is not enough. You can't switch to something you've never heard of, and most people don't even get the concept that "switching" is a thing; they see a computer as an appliance that comes a certain way and that's how it is.
What I think is true is that if the technology is good enough, like Proton works on most games and so on, then something like a Steam Machine which was not really viable years ago becomes viable. So if someone with money and marketing muscle makes some kind of push for something involving Linux, Linux will be a vehicle that can take them to the finish line. That's what makes Steamplay important.

There are alot PC gamers. Many of them are aware of Linux and that you can switch to it. Those who don't will be guided by them. Awareness about Linux will then grow alongside our marketshare.
But before anything can happen Linux must be able to run most if not all games. So important issues like EAC and BattleEye working with wine needs to be resolved.
Your points are valid. Alot people don't know about Linux or even that they can switch. But if Linux becomes even better than it currently is, then PC gamers will be first wave to get news about it and switch. That first wave of people will create a buzz, and that will further help.

But you are right, we need someone with money and marketing muscle to push Linux, when the time you can play almost all your games on it. I hope Valve will do that. Until that, the progress will be slow. I really believe people will switch. If they don't know about Linux they will through word of mouth. But the most crucial thing is that games must work. If it doesn't people will switch and not be happy and switch back. Foundation has to be solid as it is with some games running through Proton (Doom, Witcher 3, etc).


Last edited by Linuxwarper on 30 June 2019 at 9:26 pm UTC

Quoting: NeverthelessBut the point is: Proton games are not Windows only games anymore!

I agree! Steamplay is useful for to play broken or outdated Linux ports with more performance.

Quoting: Purple Library Guy

Quoting: Eike

Quoteif in short Linux was definitively better as a thing to use than for instance Windows, and specifically if you could play all the Windows games and Linux continued to kick ass in other ways, that this would inevitably result in a mass exodus from Windows and a rapid growth of Linux market share. I don't think that is actually true.

Hm... Not sure about that. It's not like I hear Windows users tell me all the time how they are just loving their Windows. And early next year, when Windows 7 is supposed to die, we'll hear that even less. There might be a barrier of "nobody's using it". If everybody knew someone who also runs Linux (and maybe could be of assistance), I could imagine this mass movement.

I'd certainly love to see that Windows 7--> Linux migration scenario materialize. But I've seen a couple of other "This MS screw-up should drive people our way" events that never worked out, so I dunno. Time will tell.

I have another machine with a legit Windows 7 SP1 pro licence that I use exclusively for gaming (Uplay and Origin) and I hope that before the Win7 EOL Proton will be 100% compatible with all the Steam games, because I'm not going to install Win 10 on that machine.

For daily stuff and alternative gaming I use this Ubuntu machine.

My new accounts since my Linux incursion.
https://steamcommunity.com/id/DigitalCHE_Reloaded

https://steamcommunity.com/id/CaS-BiGIS-ProtonicEdition

https://steamcommunity.com/id/CAS-BiGSUBiS-ProtonicEdition

https://steamcommunity.com/id/CAS-IndieS-ProtonicEdition

(And my first account from my fulltime Windows era. I don't buy games anymore with this)
https://steamcommunity.com/id/Digital_CHE

One thing is the genuine love for a platform and another thing is the fanaticism.
The zealotry for this or that platform is bad...
I am only fanatic for to play the videogames I like to play (on Windows 7 because I need it, and on Linux because I love it)

Quoting: EikeTo be precise, the amount of Linux gamers did rise - at a similar extend as the amount of Windows gamers.

You talk about decimals now, Eike. It's nothing. Nothing. You must focus on the broader picture and stop counting breadcrumbs. One percent market share is nothing, never was, never will be. Totally insignificant.

Mac is at, what, 6-7 % now? And they are barely supported more than us. And they are SEVEN TIMES larger.
When I talk about seeing changes in our market shares, I talk about changes counted in percent. Not per mille, so tiny that it becomes statistical insignificant.

When talking about market shares, we really should be talking in five percent intervals, rounded to nearest fiver. Windows is roughly 95%, Apple is roughly 5%, we are roughly 0%. That's how the decision makers see us.

I have to give it to you guys though: I kind of envy the confidence at display here. It takes SOME confidence to actually believe one has understood something that pretty much the entire multi-billion industry fails to see... ;)

Quoting: EikeYou did discuss this with the hundreds of developers doing Linux releases, didn't you?

Obviously you now are talking about indie devs. You really can not count a two-three man team and spare time single developers as the same as projects with hundreds of developers working for years. In this context they are irrelevant. They operate under a completely different cost VS gain scenario, and their projects are a lot less complex. A game is not a game on equal terms.

What we talk about here are the games that makes headlines and enter top seller lists. The big titles from the big developers. Those that do not support us today and that will not ever do that unless our share increase in a significant amount. That's where Steam Play has a role to play. Not to enable us to play yet another little indie.

Remember when the news broke about the Steam Machines and SteamOS and all that? All of a sudden we were promised a LOT of big titles. I'm sure you remember?
Now, why did that happen? It happened because they assumed that Linux would jump in market share to a level worth supporting.

That never happened - so the ports were then cancelled again. And our market share has not risen since then, not even by a mere percent.

Linuxwarper exemplified this by making a very good practical point:

Quoting: LinuxwarperThey earn more by spending the resources to make a DLC for Windows instead.

This is how market economics works. This is how a business must operate.


Last edited by Beamboom on 1 July 2019 at 10:57 am UTC

Quoting: ShmerlTotal number of Linux gamers affects sales, not market share. From what I've seen, total number of Linux gamers has been growing, year after year.

They are not two different things. It evens out since the larger platform they currently support have an even higher increase in numbers. Think about it: For the relative shares to remain the same, roughly a HUNDRED new Windows gamers must be added for every single Linux gamer. This is why the percentage number is what matters.

Supporting an entirely new platform is no small decision to make. It's not a "free" new group of potential customers.
To prepare their customer support (how many of those are well experienced with Linux, you think?), maintaining a second code branch, relate to bugs on one more platform... That is a massive decision. And you don't do that for a potential increase in sales of one percent.

I mean - why do we even discuss this? Just look at what's happened these years - or rather what's not happened. The industry is not made out of imbeciles. It's not like we sit here and have understood something that the entire industry has missed.


Last edited by Beamboom on 1 July 2019 at 10:59 am UTC

Quoting: BeamboomMac is at, what, 6-7 % now? And they are barely supported more than us. And they are SEVEN TIMES larger.

Apple's approach to gaming is quite hostile. Poor OpenGL support and have deprecated it in favor of Metal. Speaking of low level api, Vulkan isn't officially supported. Furthermore you can buy a IMac 27 inch with 5k Retina display, Radeon Pro 570x 4GB, 1TB Fusion drive and 8GB 2666 memory for $1800 or you can build a better gaming PC yourself for $945. Granted that build doesn't have a 5K display but for major games you can't expect such a Radeon Pro 570x to keep up with 5K.

I think if there were less of these issues for Mac devices, they would get more support. Also Valve is quite committed to Linux. So they will most likely capitalize on marketshare, unlike Apple who are neglecting gaming.


Last edited by Linuxwarper on 1 July 2019 at 11:44 am UTC

Quoting: LinuxwarperApple's approach to gaming is quite hostile. Poor OpenGL support and have deprecated it in favor of Metal. Speaking of low level api, Vulkan isn't officially supported.
[...]
I think if there were less of these issues for Mac devices, they would get more support. Also Valve is quite committed to Linux. So they will most likely capitalize on marketshare, unlike Apple who are neglecting gaming.

Absolutely, and I did consider this when writing my comment. But I came to the conclusion that my point still stands in regards to the market share perspective.

If we had the same market share as Apple, I am one hundred percent convinced that the support for Linux would be higher than Apple, for the exact reasons you mention.

But my main point is to try to establish an understanding on just how small we are in this pond. That perspective is fundamental when considering the pros and cons for Steam Play.

Quoting: EikeIf releasing for Windows only and losing 1% of customers, it won't hurt me and I wouldn't change anything.
If releasing for Windows only and not losing customers (because I see Linux users buying it nevertheless) - why the hell would I change anything?!?
They're already buying, why invest money?

Looks like your only problem is that companies don't invest money on the "official" Linux version of their games.

I love video games and I want to play the games that I really like. Thanks to proton I'm able to play these games on Linux just by pressing a button.

Why I have to buy games that I don't like just because they have native ports? Even if some native games are just wine wrappers and some ports run worse than the Windows version on Proton.


Last edited by aldy on 1 July 2019 at 3:05 pm UTC

Quoting: aldy

Quoting: EikeIf releasing for Windows only and losing 1% of customers, it won't hurt me and I wouldn't change anything.
If releasing for Windows only and not losing customers (because I see Linux users buying it nevertheless) - why the hell would I change anything?!?
They're already buying, why invest money?

Looks like your only problem is that companies don't invest money on the "official" Linux version of their games.

I love video games and I want to play the games that I really like. Thanks to proton I'm able to play these games on Linux just by pressing a button.

I understand that, and it's ok. It just shouldn't be presented as a means to advance native Linux gaming. If they don't care for 1% which won't buy their games if they don't port, why would they care for 0.5% which may or may nor be able to play their game under Proton?

Quoting: aldyWhy I have to buy games that I don't like just because they have native ports?

Nobody said so.
Have fun! (I mean it.)

Quoting: BeamboomI have to give it to you guys though: I kind of envy the confidence at display here. It takes SOME confidence to actually believe one has understood something that pretty much the entire multi-billion industry fails to see... ;)

Well, to be fair, the fact that the GNU/Linux desktop in general and Linux gaming in particular have gotten to where they are today is baffling under any kind of business-minded consideration. I think the ideology at our roots actually plays more of a role than we may sometimes admit. Just seeing Paradox recently saying that they will still consider porting future games to Linux on a case-by-case basis even though they have barely been turning a profit on them if at all, to me, is quite impressive. Many companies would (and have) just give up on us completely and choose easier ways to make more money (the DLC example being a very good one).

I'm not too much of an idealist, but in this case our track record shows that we can at least keep some hope alive despite slim odds :)


Last edited by Salvatos on 1 July 2019 at 3:44 pm UTC

Quoting: BeamboomBut my main point is to try to establish an understanding on just how small we are in this pond. That perspective is fundamental when considering the pros and cons for Steam Play.

Definitely we are small. Even if we do get 5% marketshare and the game in question will be supported on Stadia, companies like EA may take the lazy route and not release a native port and let Steam Play do the work for them. But even if they take that route, as long as Vulkan is available for SteamPlay for games it will still add tremendously to Linux. There is barely any difference between a native Vulkan game and one that runs through Steam Play. It's not even close to as pronounced as running game on Steam Play through DXVK.

I am optimistic about 2020. I just hope Valve can resolve anti cheat before end of year.


Last edited by Linuxwarper on 1 July 2019 at 6:06 pm UTC

Quoting: BeamboomYou talk about decimals now, Eike. It's nothing. Nothing. You must focus on the broader picture and stop counting breadcrumbs. One percent market share is nothing, never was, never will be. Totally insignificant.

[...]

What we talk about here are the games that makes headlines and enter top seller lists. The big titles from the big developers.

We won't get anywhere when we cannot even agree on what we're talking about. When I talk about the absolute number of Linux users having grown well in the last years, and you answer that the percentage is still low, this will lead nowhere. Neither when I'm talking about all games and you insist this would be only about AAA games.


Last edited by Eike on 1 July 2019 at 8:07 pm UTC

Quoting: GuestThere are other reasons for not wanting to rely on Wine for gaming, even if Vulkan is used natively (by the way, a Stadia port does not mean a Windows version will also use Vulkan - in fact, there's very little chance of that if the Windows version is using DX11 or DX12).

Ancillary benefits from natively developed games include: better tooling, more stable drivers, investment in resolving certain compatibility problems, non-reliance on Microsoft, different compilers (often can spot problems, so I would imagine a Stadia port would help here as well as native for desktop GNU/Linux), possible peripheral devices being better supported (if more native games exist, and market share grows), etc.

But of course Wine (and anything based off it), eON, dosbox, etc, all help individuals play games they want without locking themselves into Windows. So it's very important to have that.

Basically there is _plenty_ of room for both native and wrapped games, and both are absolutely needed.

It goes without saying that native ports is the ultimate outcome. But market share must be higher for that to happen. Even if we reach market share that rivals Mac, developers could be lazy and rely on Steam Play for easy money. So I expect gaming on Linux will go through a phase of many games being played through Steam Play. And if that is what we have to settle with, until market share is so significant they can't neglect us anymore, then I hope the developers at least ensure best possible compliance with Steam Play (Vulkan, no drm, no middleware etc).


Last edited by Linuxwarper on 1 July 2019 at 7:21 pm UTC

Quoting: GuestAbsolutely agreed.

I also forgot to mention another benefit of Wine in general (and I'll stick to writing "Wine" because this applies outside of Steam) and that's for archival. Old games, some even with GNU/Linux native versions, may not run properly on modern systems. Wine helps resolve that in some cases, similar to dosbox, though I daresay with increasing complexity of larger modern games this won't be as viable in the future.

...of course source code would be even better, but hey, that's orthogonal to the conversation really.

Indeed. I hope though that Steam Play can be able to achieve its goal of increasing market share before it's effort is shutdown by companies like Microsoft putting in software in games that will make the games not run as good as they do currently. That is my only concern for Steam Play. Otherwise I am optimistic that in long run (three years from now) the market share will surely have increased.


Last edited by Linuxwarper on 1 July 2019 at 8:05 pm UTC

Quoting: BeamboomThey are not two different things. It evens out since the larger platform they currently support have an even higher increase in numbers. Think about it: For the relative shares to remain the same, roughly a HUNDRED new Windows gamers must be added for every single Linux gamer. This is why the percentage number is what matters.

Supporting an entirely new platform is no small decision to make.

Related, but quite different. Let's say you have a hundred million users altogether. Even 1% out of that, already means a million users. That's a lot. I.e. probably enough to cover your production costs through sales. Who cares if total number is hundred million? For viability, it's not the market share that matters, it's the size of your market (which means total number of your users).

Or think about it this way - 90% of 10 is 9. Would it cover your costs? So forget about percentages. Absolute numbers rule.


Last edited by Shmerl on 1 July 2019 at 11:47 pm UTC

Another problem Linux gaming faces is the ROI (Return on Investment) of shipping a native Linux game. Even when shipping a native Linux game IS profitable, we often miss out on those games because the ROI is so slow. There are fixed costs for shipping native games that don't scale with the number of customers, and that can reduce the overall ROI % of shipping a native Linux game.

From the perspective of a greedy big AAA publisher with thousands of angry investors screaming at them to throw more lootboxes and mtx into their games, spending $10 to make $12 profit vs spending $10 to make $100 profit, is a no brainer, you're always going to go with the best ROI. They get a better ROI for Windows only games.

In that circumstance, it doesn't matter how high our numbers get, unless we reached something like 20 or 30% marketshare, we wouldn't those AAA games from those publishers. It doesn't matter how much we scream at them "no tux no bucks" they're quite happy to answer back: "no problem" then ignore us.

But lacking those games also keeps gamers from switching to Linux. Hence why we need Proton in the short term.

But it's too early to judge the situation, Proton has only been out for 10 months. In 10 months we have seen some positive signs, a trend of growth emerging on Steam stats, fewer people on GoL's stats indicating they're dual booting, positive youtube coverage of Linux, and Valve has shown unwavering commitment. This is all great but it's still too early to draw any conclusions.

For a start, Proton while amazing is still lacking some crucial things needed for success, like EAC/BattlEye compatibility. Multiplayer PvP games are THE big games of our present, we NEED those games on Linux and right now they're the most likely type of game to not work. Fortnite, PUBG, Rainbow Six Siege, these are huge popular games and completely unplayable on Linux right now.

We also shouldn't assume Proton would be so effective that it would draw over huge masses of gamers overnight. Never underestimate how difficult it is to ask someone to change something as crucial as OS.

People hate change, just look at how difficult it has been for Microsoft to get people off Windows 7. Windows 8 officially became available to the general public in 2012. That's 7 years ago! 7 years of bombardment of advertising, 7 years of desktop notification reminders on Windows 7 telling people to upgrade, then several years of offering Windows 10 as a free download, and even sneaky tactics like shipping Windows 10 as a free automatic update to Windows 7. For years Windows 7 hasn't been available in retail stores on laptops/desktops for years, and installing Windows 7 on a modern PC is a nightmare as drivers for modern hardware have to be sideloaded into the installer, among other countless problems. Windows 7's EOL is coming in less than 12 months.

Yet look at the Steam Stats for 2019. 1/4 of Steam users still use Windows 7. That's how much people hate change. Switching to Linux represents a HUGE change. And also a step backwards in many ways, since it means switching to an OS with less hardware support (forget customising your gaming keyboard/mouse's RGB lights and don't forget NVIDIA's crappy drivers), access to far fewer games and often worse performance.

The fact that, under all of these conditions, Linux is not only staying relevant but also seeing even a tiny silver of positive growth, is simply incredible and something to be grateful for.

We have reasons to be optimistic, but at the same time, we need to also be realistic.

Quoting: gradyvuckovicWe also shouldn't assume Proton would be so effective that it would draw over huge masses of gamers overnight. Never underestimate how difficult it is to ask someone to change something as crucial as OS.

People hate change, just look at how difficult it has been for Microsoft to get people off Windows 7. Windows 8 officially became available to the general public in 2012. That's 7 years ago! 7 years of bombardment of advertising, 7 years of desktop notification reminders on Windows 7 telling people to upgrade, then several years of offering Windows 10 as a free download, and even sneaky tactics like shipping Windows 10 as a free automatic update to Windows 7. For years Windows 7 hasn't been available in retail stores on laptops/desktops for years, and installing Windows 7 on a modern PC is a nightmare as drivers for modern hardware have to be sideloaded into the installer, among other countless problems. Windows 7's EOL is coming in less than 12 months.

Fortunately for us, in some ways, switching to Linux can be seen as a less drastic change than "upgrading" your Windows version from a UI point of view (e.g. the whole Windows 8 tiling fiasco). Nothing short of a world-changing revolution will make masses of people switch to Linux overnight, but every milestone we reach and every mistake Microsoft make results in more people willing to give it a shot, and in turn more people being happy with the change, and more people hearing about Linux, and so on. We have to be prepared for the long run because this was never something we would win in a handful of years. Fortunately for us, we already live in a time where a GNU/Linux OS is damn comfortable. And for many of us, outright better than a Windows OS, never mind the fewer video games.