I already pointed out there is a difference between user count, and marketshare.
Pointing out that the margin of error for the polling data is likely higher than the listed marketshare for linux users is not a "bias".
Pointing out another statistic that has effectively been a flatline for over a decade is not a "bias".
Asking you to post some sort of data to back up the idea that Valve's published data is somehow flawed is not a "bias", however your refusal to do so could be shown as evidence of your own bias.
Pointing out that it requires a polling error of an entire order of magnitude to make the marketshare within Valve's published statistics actually be a meaningful percentage is not a "bias".
The fact that you've actually done the childish thing of stating that you're not even going to bother reading my post that you're replying to, because I'm the one being immature? You pull some 150,000 user figure out of nowhere, based on an assumption gathered from who knows what, and you claim to have made a point? I point to statistics, and you just make up random numbers yeah... I'm pretty sure the number of people who would consider that "valid" would still likely be a higher percentage than what Valve is claiming the percentage of their users running linux happens to be.
150,000k was an example, I have to use an example as accurate user base figures direct from Valve are not available. Sorry, but I can't see the issue here. Not having accurate user base figures to bring meaning to pointless percentages is far from showing any form of bias.
You're still replying like you're on the defensive, can't you reply in a friendly manner? Fact is, developers are still releasing titles under Linux so they must see some value in the platform beyond the percentages shown in the Steam statistics. This is undoubtedly a great thing for all PC users and bringing Windows into the discussion will not change that.