Steam Hardware Surveys - are they accurate?

[HAL_404]

just had a Steam pop up window ask me to take a hardware survey so I did. Interesting thing is the OS install date it showed

 

[scojer]

Are you a time traveller?

 

[Ranulfo]

Pretty sure I've gotten the same date the last two times I've done a survery over several years.

 

[LukeTbk]

That what an empty date stamp tent to be interpreted has in UNIX timestamp language (they count with in second after midnight december 31).

 

[gamerk2]

Looks like an empty date field being converted to a UNIX timestamp to me.

 

[Scoobydo2]

I don't feel they are accurate. I had AMD video cards for 10+ years and it never prompted me to do a survey. Since buying NVIDIA, it asks me every 2 months to do a survey. I have a suspicion there is something there (paid for by Nvidia) to inflate numbers for their video cards.

 

[Armenius]

I don't feel they are accurate. I had AMD video cards for 10+ years and it never prompted me to do a survey. Since buying NVIDIA, it asks me every 2 months to do a survey. I have a suspicion there is something there (paid for by Nvidia) to inflate numbers for their video cards.

It's completely random. There are 120 million active users on Steam, so your chances of being picked are extremely low. I've had NVIDIA cards since becoming a Steam user when the beta launched in 2003 and I've only been picked 4 times in that 18 year period.

 

[Comixbooks]

Steam is broken half of my images on the app are xs.

 

[LukeTbk]

I don't feel they are accurate. I had AMD video cards for 10+ years and it never prompted me to do a survey. Since buying NVIDIA, it asks me every 2 months to do a survey. I have a suspicion there is something there (paid for by Nvidia) to inflate numbers for their video cards.

Has a reference had steam for more than 15 year's, not sure if I ever had a survey, I thought it was automatically run by it without asking.

In the last 6 year's:


NVidia discrete gpu market share seem to stay between 70 to 80% while amd high 10x to 30%

Looking only at DX12 Gpus on the hardware survey, seem to show 70% Nvidia to 10% AMD, 6.12% intel.

So of the discrete DX12 GPU market that match the Nvidia domination (2014-2015+) era, that seem to be 87.5% Nvidia to 12.5% AMD which seem to low versus the Dgpu market share over that period.

Is the market where steam is popular different than the world market ? Is gamers more Nvidia heavy than the population has a whole (not sure how big and relevant on gamer that still buy dGPU in that intel dominated era), is the personality of the AMD buyers a bit different and less likely to answer survey or is there like you said some paid for bias, not sure how much people use that survey in their buying decision and I do not remember ever popping up in any marketing material on the newegg of the world either. Maybe a lot more of the model not making the cut to show up are relevant AMD model (5500xt-5600xt) etc... than Nvidia one and that add a bit to the discrepancy.

For example more than 30% of steam userbase seem to be chinese now, if that hardware market is different than ours that would make steam numbers look strange to us even if they are perfectly legit just by how giant the China part of the survey could be.

 

[Armenius]

Had a reference had steam for more than 15 year's, not sure if I ever had a survey, I thought it was automatically run by it without asking.

In the last 6 year's:
View attachment 374944

NVidia discrete gpu market share seem to stay between 70 to 80% while amd high 10x to 30%

Looking only at DX12 Gpus on the hardware survey, seem to show 70% Nvidia to 10% AMD, 6.12% intel.

So of the discrete DX12 GPU market that match the Nvidia domination (2014-2015+) era, that seem to be 87.5% Nvidia to 12.5% AMD which seem to low versus the Dgpu market share over that period.

Is the market where steam is popular different than the world market ? Is gamers more Nvidia heavy than the population has a whole (not sure how big and relevant on gamer that still buy dGPU in that intel dominated era), is the personality of the AMD buyers a bit different and less likely to answer survey or is there like you said some paid for bias, not sure how much people use that survey in their buying decision and I do not remember ever popping up in any marketing material on the newegg of the world either. Maybe a lot more of the model not making the cut to show up are relevant AMD model (5500xt-5600xt) etc... than Nvidia one and that add a bit to the discrepancy.

For example more than 30% of steam userbase seem to be chinese now, if that hardware market is different than ours that would make steam numbers look strange to us even if they are percently legit by how giant it is.

Steam has a version for the Chinese market now. I wonder if they separate survey results or will?

 

[3dprophet]

Steam has a version for the Chinese market now. I wonder if they separate survey results or will?

I'd be surprised if the Chinese version isn't a completely separate ecosystem.

 

[Deleted member 83233]

I get surveyed all the time. The problem is, I have Steam on around ten machines. It always seems to survey me when I’m logged in on a laptop or something, not when I’m on my main game PC, so the results they get from me always look lower end

 

[SOAREVERSOR]

It's fine it's just a bit random rather than slamming everyone each month which would gain better results. AMD has always been a largely non issue hardware side though. Reality is most people game on laptops and 1080p 60hz screens and get a worse than console experience but they pirate games and use sales. Hence the turbo inflated i3/i5 and 1050 gpu type items. We here are just a fucking blip in their stats.

Also who's run a cyber care, raise hands (me raises hand halfway). A few of my friends ran them in the USA and I helped support them. That gaming side largely mimics the corporate side where AMD is just a no no because of bugs. I don't just mean drivers I also mean chipset and controller issues, people are just gun shy.

I get surveyed all the time. The problem is, I have Steam on around ten machines. It always seems to survey me when I’m logged in on a laptop or something, not when I’m on my main game PC, so the results they get from me always look lower end

I dunno it hits me on my desktop all the time and I have server grade SSDs so oh well.

 

[FrgMstr]

I have heard a lot of behind the scenes conversations about this over the years, and the general consensus is that the numbers are fubar enough that Valve is sorry they ever started making the data public, but were concerned with the public perception of pulling it back.

 

[Blackstone]

You know what else is is FUBAR? That Subway will give Gaben triple meat on his subs but when I try to do thing I am told double meat is the limit.

 

[vlad_8011]

He'll, it's not accurate. Not to mention people who is not taking part in the survey (clicking "cancel"), or simply noting is popping out on their machines, but I found very tricky way of how this works.

Officially this survey pop out random, but I don't believe in miracles, and that survey NEVER ever popęd out when I had AMD GPU - not a single time, but 2 times when I had GTX 1060, 3 times when I was on GTX 1080Ti and 1 time on RTX 3080. Is this coincidence? Beacuse I don't thing so.

 

[polonyc2]

it's accurate...the vast majority of PC gamers do not own the highest end hardware...it's more about value which is why the Nvidia 1060 is the most popular card

 

[vlad_8011]

It WOULD be accurate if they would release info, that for one week they will make survey, and send it to all users.

 

[Comixbooks]

https://www.pcworld.com/article/394207/how-to-make-your-pc-take-the-steam-hardware-survey.html

Just found this on how to take the survey without the popup that only comes up like twice a year.

 

[vlad_8011]

https://www.pcworld.com/article/394207/how-to-make-your-pc-take-the-steam-hardware-survey.html

Just found this on how to take the survey without the popup that only comes up like twice a year.

Yeah, I tried it before, but there were game-oriented surveys (and VR) and none of them actually send any info - these were stuck with sending information page.

 

[HumanBias]

Has a reference had steam for more than 15 year's, not sure if I ever had a survey, I thought it was automatically run by it without asking.

In the last 6 year's:
View attachment 374944

NVidia discrete gpu market share seem to stay between 70 to 80% while amd high 10x to 30%

Looking only at DX12 Gpus on the hardware survey, seem to show 70% Nvidia to 10% AMD, 6.12% intel.

So of the discrete DX12 GPU market that match the Nvidia domination (2014-2015+) era, that seem to be 87.5% Nvidia to 12.5% AMD which seem to low versus the Dgpu market share over that period.

Is the market where steam is popular different than the world market ? Is gamers more Nvidia heavy than the population has a whole (not sure how big and relevant on gamer that still buy dGPU in that intel dominated era), is the personality of the AMD buyers a bit different and less likely to answer survey or is there like you said some paid for bias, not sure how much people use that survey in their buying decision and I do not remember ever popping up in any marketing material on the newegg of the world either. Maybe a lot more of the model not making the cut to show up are relevant AMD model (5500xt-5600xt) etc... than Nvidia one and that add a bit to the discrepancy.

For example more than 30% of steam userbase seem to be chinese now, if that hardware market is different than ours that would make steam numbers look strange to us even if they are perfectly legit just by how giant the China part of the survey could be.

That data point is years old, here is the latest datapoint:

 

[LukeTbk]

That data point is years old

So is the message you quoted

 

[HumanBias]

So is the message you quoted

The message was from 2021, with a graph from 2019