The Steam Hardware Survey For May Is Here, It’s Not All Doom & Gloom

I guess will install Windows then...or better, maybe let's just not wait for miracle in Steam survey data :)

I care less and less. All I care is good ports and people voting with their wallets.

Steam usage must be growing so fast that the Linux adoption rate just cannot keep up. Linux still has more users using Steam than before. Unfortunately, so do Windows and Mac OS.

Arf. It's frustrating.

I hope rumours are not true that Newell just came up with Linux/SteamOS/SteamMachines to put pressure on Microsoft.
As Win10 seems like it will get good reception and "lacks" all the stuff Newell was afraid of, some people argue Valve might abandon Linux sooner or later. After all, SteamMachines missed the fantastic "Win8" time window to take off - and that sounds reasonable to me. :/

I wish Steam had a statistic of the sum of people running it on the various platforms at any given time, and how many of them are playing games.

For me, the survey only seems to happen when I am installing steam, and I usually just skip it, and I am pretty sure so will a plenty of other Linux users.

I hope rumours are not true that Newell just came up with Linux/SteamOS/SteamMachines to put pressure on Microsoft.
As Win10 seems like it will get good reception and "lacks" all the stuff Newell was afraid of, some people argue Valve might abandon Linux sooner or later. After all, SteamMachines missed the fantastic "Win8" time window to take off - and that sounds reasonable to me. :/

I kind of doubt Valve would just ditch it due to their investment in Vulkan.

And as long as Microsoft has it's own app store, they are still a monetary threat to Valve, and afaik Windows 10 still has an (albeit optional) app store.

I doubt that Steam will switch over using Windows for their Steam box. The effort they would have done until now would be pointless. And the license fee, which they would have to pay to Microsoft is surely high enough.

I hope rumours are not true that Newell just came up with Linux/SteamOS/SteamMachines to put pressure on Microsoft.
As Win10 seems like it will get good reception and "lacks" all the stuff Newell was afraid of, some people argue Valve might abandon Linux sooner or later. After all, SteamMachines missed the fantastic "Win8" time window to take off - and that sounds reasonable to me. :/

Microsoft hasn't given up walled garden approach. It's still there and Valve knows it very well. Fool me once - shame on you, Fool me twice - shame on me.

Linux users really should stop being afraid of Windows. Seriously, all I hear every time when Microsoft releases new Windows is how it will finally destroy Linux desktop. That's amazing PR pull if you ask me. Windows still costs 100USD/EUR at retail. Microsoft will offer similar discounts to illegal users, but for very short time and they are similar to Windows 8. DirectX 12 doesn't mean all games will magically run faster - you have to use low level functionality for that. DirectX12 being locked Windows 10 only will limit it's addoptation.

And so on and so forth.

Also what it got to do with blip in statistics I wonder. Yes, there are hopes and fears, but Linux users should be better than that.

It's only logical. There are too few AAA releases, even fewer first day releases and virtually no stable such. The Linux triple A game market consists of 3-4 aging game ports (quite buggy upon release) and a lot of vague promises.

For anybody that might be worried about the decline I'd like to point out that I've gone years without receiving a Steam survey, only to turn around and have three or four in a year. This year for example I've received four surveys. One on Windows and three on Linux, two of which were in the same month a couple days apart, so clearly Valve's polling method is far from perfect.

We have no idea how they are sampling their data, and to state this again, it’s a survey.

They should introduce a survey day eg. the 15th of every month to make the survey representative. It also wouldn't be a problem at all to just send 1 additional byte to identify the OS every time users log into Steam. Maybe this is against Valve's user data policy.

Steam usage must be growing so fast that the Linux adoption rate just cannot keep up.

Haven' t heard of a person that used linux and said "I want to use Win to play game XY". For most non-Linux games there should be good Linux alternative titles. Except racing games but there will be GRID soon.

For me, the survey only seems to happen when I am installing steam, and I usually just skip it, and I am pretty sure so will a plenty of other Linux users.

Maybe Windows users tend to reinstall their OS more frequently than Linux users.

We have no idea how they are sampling their data, and to state this again, it’s a survey.

They should introduce a survey day eg. the 15th of every month to make the survey representative. It also wouldn't be a problem at all to just send 1 additional byte to identify the OS every time users log into Steam. Maybe this is against Valve's user data policy.

ehem...

...this is intentional on valve's side... I dont know the reasons... but they already know exact number of users per platform and the exact amount of time and money... with no room for interpretations...

Why arent they releasing this data...? well, my guess is that knowledge is power and this way only they have this knowledge...

What is sad is that mac users triple us in numbers...

it is also true that they have almost double the number of games... and probably about 3 or 4 times more AAA games than linux... but still...

What is sad is that mac users triple us in numbers...

The stats from Steam are in clear correlation to the general stats for desktop usage we find, so none of these numbers should surprise anyone.

Netmarketshare.com reports:
Linux: 1.5%, Mac OSX10.10: 4.2, "others" 3% (mostly other macs I assume) and the rest are Windows flavours.
Source

Wikipedia refers to Net Applications results:
Linux: 1.5%, Mac OSX (all versions): 7.3, the rest is Windows.
Source

I could go on. Bottom line is that there's no reason to try to explain the numbers from Steam in any other way than that it really reflects the reality out there. This is how it is today. There's absolutely no reason to look for excuses for why these numbers could be wrong in any major way.

Participated in the survey with my Linux machine yeah ;-)

It's only logical. There are too few AAA releases, even fewer first day releases and virtually no stable such. The Linux triple A game market consists of 3-4 aging game ports (quite buggy upon release) and a lot of vague promises.

SteamOS would definitely require games like just announced Fallout 4. It will take long time to get gamers interested about SteamOS.

They should introduce a survey day eg. the 15th of every month to make the survey representative. It also wouldn't be a problem at all to just send 1 additional byte to identify the OS every time users log into Steam. Maybe this is against Valve's user data policy.

Your first point: they do this already. The surveys go out at the same time.

Your second point: they don't need to. They already know how many people are loading steam what each OS. This information is going to be accurate because they have a 100% check of it (client identify at login). The point of the survey is to maybe find out what specific OS you are running (Ubuntu 15.04 64-bit, etc), but more likely what hardware you have on-board. You already know you see platform specific sales and notices, of course they know.

0.88% (+/- error) is the real number, accept it. By the way if you convert the percentages to the numbers of users the Linux user base is growing (roughly 30k, a month). The total steam user base is growing two orders of magnitude faster though (3-4million per month).

Also, if you take a look at the numbers in more detail you can see that Linux isn't growing in the same areas that Windows is. Easiest way to pull something from the survey data is to look at languages. People generally don't change the language steam is set to, and language correlates somewhat with country.

Linux (two fastest growing):
French 2.14% +0.16%
Spanish 1.88% +0.22%

Windows (two fastest growing):
Thai 2.53% +0.99%
French 3.60% +0.18%

Mac (two fastest growing):
Spanish 2.91% +0.17%
Dutch 0.83% +0.13%

I've no idea what is happening in Thailand, but that's a change equal to 3 million new or returning users there.

With so many different marketshare reporting sites, who should we believe? They all seem to have different numbers ranging from less than 1% up to over 5%. So who is accurate and who is not?

Id like to see a company that just makes amazing AAA games for Linux / SteamOS only, and watch the other platforms squirm. haha