No, these stats are not increasingly becoming less reliable and skewed by the use of virtualization software like Parallels and Boot Camp, because what these numbers try to evaluate is OS usage, not hardware usage, so it doesn’t matter if you use Vista (or any other OS) through virtualization or natively instead, what matters is that you’re using it, it doesn’t matter how.

Apparently Tom Krazit is claiming that Net Applications is claiming that Windows Vista is eroding Mac OS X sales, based on the fact that Net Applications simply (correctly) concluded that Windows Vista is selling more than Mac OS + MacIntel and surpassing Mac OS + MacIntel market share after analyzing numbers while never claiming that was in prejudice of Mac OS + MacIntel market share.

“The problem with the data, according to Krazit, is that Net Applications only monitors 400,000 websites, most of which are small/medium business sites and that the vast majority of computers sold come with Vista preloaded.”

So let me see if I get this straight: the problem to Krazit here is that the computers come with Vista preloaded. Because if they would come with Mac OS X preloaded then I guess it would be no problem, is it? OK, I think I got it then: basically, what he’s saying is that the problem here is that people and businesses choose a PC with Vista instead of a Mac with Mac OS and that makes comprisons unfair, because a comparison is only fair if it concludes something negative to Vista and positive to Macs. Got it.

“Net Applications claimed that Vista usage shot up in February, right about the time that Vista first started shipping on most new computers sold.”

When Vista usage provenly shots up from 0.93% market share in February to more than 4.52% now, I think you don’t call it “claim” anymore, you call it… something like “fact” I think.

“So the fact that Vista went from nothing to something would make the numbers look a lot more impressive.”

Too bad they really are impressive because these numbers are facts and against facts there are no arguments and Vista did go from nothing to “something” (”something” bigger than Mac OS + MacIntel market share in all these years, I should add).

“Krazit did a little of his own research and found the opposite of what Net Applications is reporting to be true.”

I did a little of my own research too and found the opposite of what Net Applications is reporting (and what is Net Applications reporting? That Vista is selling more than Mac OS) to be false.

“Let’s look at the overall trends since last summer. The combination of Windows PCs (Win 2000, XP and Vista) visiting Hitslink sites in July 2006 accounted for 90.39 percent of traffic. Of course, there was no Vista then. This June, the same combination accounted for 90.46 percent of traffic, implying that if Vista is eating into anything, it’s Windows XP and Windows 2000 usage.”

This just in: Earth is round and CrunchGear is biased.

“The same comparison applied to Mac OS X? Last year, the combination of Intel-based Macs and older Macs accounted for 4.29 percent of traffic to Hitslink’s sites. This June, it was 6 percent of traffic, with Intel-based Macs growing as PowerPC Macs decline.”

What sense does it make to compare today’s market share with the market share of one year ago, instead of with the market of when Vista was released to the general public, as the Net Applications conclusions are about Vista and not about Windows in general? Could it be because, if one would do that, one would conclude Vista’s market share grew almost 5% and Intel-based Macs and older Macs combined actually slipped?

“So compared with last year, Mac OS usage is increasing faster than Windows usage.”

So compared with February 2007 (which is when Vista was released to the general public, which, by itself, is what we are talking about here), Mac OS usage is decreasing while Vista (which is what we are talking about here) is increasing fast.

All in all, an interesting look at how easily facts and figures can be skewed indeed.