Support us on Patreon
PayPal
The latest Steam Hardware Survey is now available and if you take it at face value it doesn't look so good for Linux. I am remaining positive as always about it. I'm not going to give it some sensationalist headline about it being terrible or anything like that, I will simply give you some thoughts on it.
First up, the facts:
Linux is currently at 0.84%, which doesn't sound like a lot and there is a drop over last months 0.90%.
Now, onto some thoughts:
The key thing to remember is Steam overall is always growing, so a lower overall percentage of Linux users doesn't necessarily mean there are less Linux users on Steam (it could actually be more, but dwarfed by also having even more Windows users on Steam). An older article from Cheese states this:
Before reading this next part, I should remind you this is an editorial. These are my personal thoughts on it and I could be wrong. I accept that, but I think it's interesting anyway.
You can make it appear by simply having different hardware or a different operating system. It seems to detect when you change things, as if it knows it needs to check on you again. This is by design of course, as the entire point of it is to show what people are currently using, so if you've changed something it wants to know about it and send it along. This is one reason why people keep saying they see it when they boot into Windows after not using it for a while, of course you will, that's a change in your setup. This is another reason why I dislike it, as that can create an unintentional bias in the results. This bias isn't against Linux though, as it would work the same the opposite way around of course. This is why I feel the results were actually a lot higher for Linux initially, as it did a survey for a big bunch of Windows/Mac users trying it and submitting it on Linux before moving back to Windows/Mac.
A good bit of reading was a recent editorial titled "A different approach to calculating the popularity of Linux gaming on Steam" which will help put your mind at ease.
I would also like to point out that I personally feel things are still looking good, as I wrote in a recent editorial.
Also remember, we have our own statistics page which is currently in BETA. Would love more feedback on that, seems people like it so far! Article taken from GamingOnLinux.com.
First up, the facts:
Linux is currently at 0.84%, which doesn't sound like a lot and there is a drop over last months 0.90%.
Now, onto some thoughts:
The key thing to remember is Steam overall is always growing, so a lower overall percentage of Linux users doesn't necessarily mean there are less Linux users on Steam (it could actually be more, but dwarfed by also having even more Windows users on Steam). An older article from Cheese states this:
QuoteFirst up, it can be easy to miss that percentages are ratios. When looking at changes over time, an increase or decrease in percentages doesn't necessarily mean that there is a corresponding increase or decrease in the population represented by that percentage.
Before reading this next part, I should remind you this is an editorial. These are my personal thoughts on it and I could be wrong. I accept that, but I think it's interesting anyway.
You can make it appear by simply having different hardware or a different operating system. It seems to detect when you change things, as if it knows it needs to check on you again. This is by design of course, as the entire point of it is to show what people are currently using, so if you've changed something it wants to know about it and send it along. This is one reason why people keep saying they see it when they boot into Windows after not using it for a while, of course you will, that's a change in your setup. This is another reason why I dislike it, as that can create an unintentional bias in the results. This bias isn't against Linux though, as it would work the same the opposite way around of course. This is why I feel the results were actually a lot higher for Linux initially, as it did a survey for a big bunch of Windows/Mac users trying it and submitting it on Linux before moving back to Windows/Mac.
A good bit of reading was a recent editorial titled "A different approach to calculating the popularity of Linux gaming on Steam" which will help put your mind at ease.
I would also like to point out that I personally feel things are still looking good, as I wrote in a recent editorial.
Also remember, we have our own statistics page which is currently in BETA. Would love more feedback on that, seems people like it so far! Article taken from GamingOnLinux.com.
0 Likes
Some you may have missed, popular articles from the last month:
The comments on this article are closed.
Like many others, I am also only seeing surveys regularly the odd times I use Windows, whereas on my 4 Linux machines that I use all the time I almost never see a Steam survey. I assume, as somebody else mentioned this is a corporate decision by Valve, as it has been going on since the Steam-for-Linux service was launched. If it is for some marketing reasons, then I think it is rather disrespectful of their Linux users.
0 Likes
autonomouse Jun 2, 2016
According to a chart in the comments section in Phoronix (https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/875653-valve-s-steam-survey-shows-linux-gaming-fall-to-one-of-the-lowest-levels-ever/page5) in February 2013, when our market share was ~2%, there were 7million active Steam users. Now, when we're at 0.84%, there are an estimated 18.5million users.
If those numbers are correct - and I haven't checked them, so please find an original source before anybody quotes this - then in 2013 there were 140,000 linux users and now there are 155,400 users.
So the userbase has gone up (by 15,400 users!), but it just hasn't kept in line with the number of Windows users, hence the lower percentage.
(7,000,000/100%) * 2.0% = 140,000
(18,500,000/100%) * 0.84% = 155,400
I haven't the time now to dig around to find out if those figures for the total concurrent number of steam users are correct, but even if they're not, please do remember that this is all relative (as is mentioned in the article).
Actually, maybe in future when referring to the steam survey, the headlines should quote the number of users, calculated in this way, rather then the percentage?
If those numbers are correct - and I haven't checked them, so please find an original source before anybody quotes this - then in 2013 there were 140,000 linux users and now there are 155,400 users.
So the userbase has gone up (by 15,400 users!), but it just hasn't kept in line with the number of Windows users, hence the lower percentage.
(7,000,000/100%) * 2.0% = 140,000
(18,500,000/100%) * 0.84% = 155,400
I haven't the time now to dig around to find out if those figures for the total concurrent number of steam users are correct, but even if they're not, please do remember that this is all relative (as is mentioned in the article).
Actually, maybe in future when referring to the steam survey, the headlines should quote the number of users, calculated in this way, rather then the percentage?
1 Likes, Who?
Mountain Man Jun 2, 2016
Quoting: ShmerlI think people should stop paying attention to this survey. It's not representative of the whole picture even for Steam itself.I agree, the results are very misleading. Unfortunately, a lot of people do pay attention to it, including developers who sometimes cite the survey results as a reason to not support Linux.
I wish Valve would release their internal statistics some time because you know they have an exact count of how many people are using which operating system and how many people are dual booting (a practice I gave up months ago).
3 Likes, Who?
Mountain Man Jun 2, 2016
Quoting: autonomouseAccording to a chart in the comments section in Phoronix (https://www.phoronix.com/forums/forum/phoronix/latest-phoronix-articles/875653-valve-s-steam-survey-shows-linux-gaming-fall-to-one-of-the-lowest-levels-ever/page5) in February 2013, when our market share was ~2%, there were 7million active Steam users. Now, when we're at 0.84%, there are an estimated 18.5million users.Your math isn't quite right. That should be 1,400,000 and 1,554,000, so the increase in Linux users was over 150,000, so even as flawed as the survey is, it's still showing steady growth for Linux.
If those numbers are correct - and I haven't checked them, so please find an original source before anybody quotes this - then in 2013 there were 140,000 linux users and now there are 155,400 users.
So the userbase has gone up (by 15,400 users!), but it just hasn't kept in line with the number of Windows users, hence the lower percentage.
(7,000,000/100%) * 2.0% = 140,000
(18,500,000/100%) * 0.84% = 155,400
I haven't the time now to dig around to find out if those figures for the total concurrent number of steam users are correct, but even if they're not, please do remember that this is all relative (as is mentioned in the article).
Actually, maybe in future when referring to the steam survey, the headlines should quote the number of users, calculated in this way, rather then the percentage?
Last edited by Mountain Man on 2 June 2016 at 3:49 pm UTC
0 Likes
autonomouse Jun 2, 2016
Quoting: Mountain ManYour math isn't quite right. That should be 1,400,000 and 1,554,000, so the increase in Linux users was over 150,000, so even as flawed as the survey is, it's still showing steady growth for Linux.
Maths isn't exactly my thing, but unfortunately I think I'm right on this one: your figures would mean that 1.4 million is 2% of 7 million. 1,400,000 would be more like 14%.
Bear in mind that this is referring to the number of concurrent users on steam - not the number of accounts. I probably should have made that clearer. I'll have a look around for an original source for the total number of Steam users, and that will give us our true number of linux users, which may well be > 1 million.
I don't know if anyone trusts those though, as the accounts may be inactive, or one person may have set up several accoutns, or whatever.
1 Likes, Who?
Mountain Man Jun 2, 2016
Oops... you're right. I multiplied the percentages by a factor of 10.
0 Likes
I dunno. No matter the this and the that and the overall growth of Steam and the survey methodology and so on, the trend has not been great lately.
I mean, we don't know for sure but I would tend to assume that however the dang survey works, it hasn't been changing over time to be progressively worse for Linux. So even if our numbers should be higher overall, a drop is a drop. And OK, maybe Steam is growing, but a drop still means if Steam is growing, we're not.
We need the Steam Machine to do well, and we need it bad. Crossing my fingers for a renewed launch with better features this Christmas.
Thing is, I really don't understand why Linux should be dropping as a proportion of PC desktops (which is the vague implication I get from the gradual drop of Linux on the Steam survey--I realize there's a lot of caveats between the one thing and the other, but all else equal that's what you'd expect that to mean). I mean, in terms of how usable and likeable the Linux desktop and general software ecosystem (including games) is compared to Windows, we're in about the best shape we've ever been except possibly for a brief moment when everyone was ragging on Vista. And we didn't have games then. Nobody's marketing Linux on the desktop, sure, but nobody ever really was. Linux off the desktop--everywhere else in computing from tiny to huge--just keeps going from strength to 500-lb-gorilla dominance.
So what's the deal?
Last edited by Purple Library Guy on 2 June 2016 at 6:28 pm UTC
I mean, we don't know for sure but I would tend to assume that however the dang survey works, it hasn't been changing over time to be progressively worse for Linux. So even if our numbers should be higher overall, a drop is a drop. And OK, maybe Steam is growing, but a drop still means if Steam is growing, we're not.
We need the Steam Machine to do well, and we need it bad. Crossing my fingers for a renewed launch with better features this Christmas.
Thing is, I really don't understand why Linux should be dropping as a proportion of PC desktops (which is the vague implication I get from the gradual drop of Linux on the Steam survey--I realize there's a lot of caveats between the one thing and the other, but all else equal that's what you'd expect that to mean). I mean, in terms of how usable and likeable the Linux desktop and general software ecosystem (including games) is compared to Windows, we're in about the best shape we've ever been except possibly for a brief moment when everyone was ragging on Vista. And we didn't have games then. Nobody's marketing Linux on the desktop, sure, but nobody ever really was. Linux off the desktop--everywhere else in computing from tiny to huge--just keeps going from strength to 500-lb-gorilla dominance.
So what's the deal?
Last edited by Purple Library Guy on 2 June 2016 at 6:28 pm UTC
0 Likes
Perkeleen_Vittupää Jun 2, 2016
So even The Linux Gamer hadn't gotten the survey for YEARS!? There's something very peculiar about this damn survey without a question.
0 Likes
See more from me