Glynor, out of curiosity whose numbers are you looking at?
There is definitely "disagreement" among the different usage share analyticics stats out there (and always has been).
The one that is showing that Windows 10 has tied OSX usage shares in August and surpassed them in September was from Net Marketwatch. This was reported all
over the place, but the news stories all basically referenced the same report. Net Marketwatch does significant "weighting" to their numbers, particularly for geography, but also for device type. That gives me a bit of pause (like the "likely voters" polls showing Romney winning in 2008, for example), and historically, that firm has been pretty optimistic on Windows Phone and IE usage shares while many others showed... Other results.
Statcounter, which has a much larger network and publishes more raw data, still shows OSX at ~8.7% usage share (depending on the particular metric used) and that, since an impressive early rise (people like free stuff), Windows 10 has stalled out at ~4.88%. They've also historically favored Chrome in browser stats. But they also have a substantially different methodology.
There were also reports on the Steam usage share numbers showing Windows 10 with huge numbers, but I think it is fairly obvious that Steam would have an overall Windows slant. While Steam absolutely has dramatically improved the gaming situation on Macs, even on Steam, only a fraction of the games are even available for the platform. OSX, for better or worse, is still not a top-tier gaming platform.
The numbers from both of the two firms don't come anywhere near close to agreeing with one another, and they also tend not to match actual shipped devices figures actually released by companies. I've seen various arguments for both companies methodologies. I'm not enough of a statistician to evaluate their accuracy myself, though most of the analysis I've seen (plus a simple reading of their published methodologies) leads me to believe that
neither is particularly accurate. My gut says that StatCounter under-estimates
both OSX and Windows 10 (and all other "also rans" in comparison to Windows 7), and skews towards heavy web users, verses machines-at-large. Net marketshare skews towards Asia, and then adds some "geographical balancing" magic sauce, and then spits out numbers. My gut says they tend to under-report OSX (a hunk of which end up in "Other" to a degree).
In any case, they're both not really measures of machines
in use, but machines in use
for web browsing (a particular set of websites). And, none of them track machines well that run ad blockers, or who have browsers that lie about their user agents and other details (so, for example, none of my machines would count correctly, and a huge number of both our Macs and Windows boxes at work would never show up in
either metric).
So, I shouldn't have said it "absolutely" above. My bad. Sorry.
One of the two companies that publicly generates stats on that shows Windows 10 nudging ahead of OSX this past month, the other shows OSX with a ~3.5-4 point lead. Both of them show Windows 7 still substantially ahead of both combined.
In my defense, I wrote it at work and I was rushed. I do know about the two major browser based trackers, and their differences. I suspect they're roughly tied, with OSX still a point or two ahead in installed base. What the two sets of numbers do agree on, though, is that Windows 10 adoption seems to have stalled a bit. It roared out of the gate, certainly (and, to be clear, I
like it). But, it isn't rapidly
taking over Windows 7, despite being both free and "adware advertised" on every copy of Windows 7 and 8 out there.
As adoption rates on Macs themselves (and iOS) have shown, making it free and easy to obtain
can move huge masses of people to perform updates. Windows has (at least so far) remained at least somewhat resistant to those sweeping transitions (despite these changes).
Author
Topic: Idle Usage Share Jibber-Jabber (Read 2404 times)