Valve Publishes Steam Survey Numbers For April 2024

[mSparks]

Not in the Linux world - which you would see if looking at the stats.

And you are surprised that something that doesn't work by default on Linux doesn't get used on Linux?

Really?

[mSparks]

Sure, arch and fedora are going to switch to X11. It will be right after they switch to 32 bit and to Ext2 as their default filesystem.

They will just continue to push gamers away as hard as they can I'm sure.

Personally, I'm going keep going with

https://youtu.be/RE5iDfBHZNQ

as long as I can. But they are doing everything they can to break it, so I don't know how much longer that lasts. .

But you just keep enjoying minesweeper and solitaire on wayland wondering why no one gives a shit.

[mrg666]

Maybe number of Linux gamers are not growing but software (Mesa, Wine, Wayland, ...) is improving nicely. I notice big improvements in performance and quality on latest FC40 using Plasma 6 and Wayland, and of course, AMD GPU. Intel is also adding a new GPU driver which could be ready soon in Linux 6.9, can't wait to try with my A770. Nvidia is a different story but I think they will change their act against Linux.

[blackiwid]

Can somebody explain to me how this numbers are supposed to be 0,1% exact? If I see voting polls the different companies have differences of 3-4% in absolute numbers and they try to calculate out if people don't admit to vote for a certain party that is controversial as example.

I recently clicked away a steam survey, but also if I would buy a steam deck I would even much more likely click it away, it's one thing to fill it with mouse and keyboard and another thing to use such controller thing.

Or does it count the OS even if you click it away automatically? I don't think they do a representative poll, like if there are 30% very young people ask 30% such people because I don't think they even have that data. So how on earth could that be 0,1% exact?

[mSparks]

Can somebody explain to me how this numbers are supposed to be 0,1% exact? If I see voting polls the different companies have differences of 3-4% in absolute numbers and they try to calculate out if people don't admit to vote for a certain party that is controversial as example.

I recently clicked away a steam survey, but also if I would buy a steam deck I would even much more likely click it away, it's one thing to fill it with mouse and keyboard and another thing to use such controller thing.

Or does it count the OS even if you click it away automatically? I don't think they do a representative poll, like if there are 30% very young people ask 30% such people because I don't think they even have that data. So how on earth could that be 0,1% exact?

steam have a 100% sample rate of steam users. some 300 million people.

0.1% is still around 300,000 less people using linux for gaming. That is not because it got better.

[blackiwid]

steam have a 100% sample rate of steam users. some 300 million people.

0.1% is still around 300,000 less people using linux for gaming. That is not because it got better.

It doesn't matter how many users they have we have polls about big countries with many millions of users, too. But they don't poll all this 300mio people so this number says nothing about how good the poll is.

Again polling of voting predictions differ from different institutes 3-4% why would Steam be correct to 0,1%? Maybe you misread my statement, I didn't say that if really 0,1% less people would use linux that this was not a lot, my question is how could they measure it that exactly?

Let's say >40 year olds click away the poll 200% than <40 year olds, because they just randomly ask people they would then not replace this poll with another >40 year old, like such demoscopes do with the voting question. So this bias that >40 year olds click more likely away such polls would not be corrected, so you suddenly would have young people in the poll over represented, now if young people would use more linux or more windows you would have also counted that OS wrong.

Now you could argue that it at least measures a tendency even i the total numbers are of by huge amount maybe, but again 0,1% sounds like no demoscopes could be that exact. Nobody advertises that their voting polls have an error of only 0.1% nobody even lists the numbers after the point.

My question would be if I get a poll and I click it away, does it count automatically my OS anyway or not, if it did that would be at least slightly better and I would trust the numbers slightly more if not then forget about it it becomes a total joke.

Again also it could be that steam deck users because of the form factors click it away more often than desktop users, which of course would hurt linux numbers.

[andre30correia]

steam have a 100% sample rate of steam users. some 300 million people.

0.1% is still around 300,000 less people using linux for gaming. That is not because it got better.

you use linux mint , and nvidia and intel? I cant understand people who continues to use nvidia for gaming in linux, better to use windows, is common sense the nvidia drivers have some problems with new distros, change to amd and the things will start to work or switch to windows if you use nvidia, even me change to red team some time ago because it works better in desktop linux

[mSparks]

you use linux mint , and nvidia and intel? I cant understand people who continues to use nvidia for gaming in linux, better to use windows, is common sense the nvidia drivers have some problems with new distros, change to amd and the things will start to work or switch to windows if you use nvidia, even me change to red team some time ago because it works better in desktop linux

nvidia has no problems with linux as long as you ignore the bullshit. in fact it is orders of magnitude faster than windows for the AI stuff that pays for me to fly helicopters for my spare time.

[mSparks]

It doesn't matter how many users they have we have polls about big countries with many millions of users, too. But they don't poll all this 300mio people so this number says nothing about how good the poll is.

steam collects its data in real time from every user on the platform.

https://steamcharts.com/app/730

The survey just adds data they cannot collect automatically by law (like what distribution of linux/version of windows was being used or what model cpu/gpu is being used)

[qarium]

it runs.. on amd (not nvidia at all, which is some 80% of gamers), but the input lag, lack of steamvr support on gnome, and extremely common crashing of the compositor are well established and more than enough reason for most to go straight back to windows - its why linux mint is so popular for steam gamers - it defaults to X11.
when a popular recommended distro like arch or fedora switch to X11, that 1.9% will easily hit double digits.
unfortunately e.g. fedora do everything possible to keep steam gamers away, intentionally hard to install the nvidia drivers, now remove X11 from the default install completely so not even AMD guys and gal gamers will use it.

I run Fedora 40 and i use Steam to run games ... and of course i run it on wayland...

and of course i am on AMD hardware and no serious open-source supporting person should support nvidia.