backtop


Print 38 comment(s) - last by inighthawki.. on May 14 at 10:02 PM

Microsoft won't say exactly how many windows 8 machines are being used

Not too long ago Microsoft was bragging that it had sold 100 million licenses for Windows 8. While the software giant was happy to say how many licenses that have sold, it's not offering any details on exactly how many of those licenses have been activated and how many copies of Windows 8 are in use. This is a question analytics firm Moor Insights & Strategy recently tackled.

Analyst Patrick Moorhead stated, "The challenge is figuring out what that [100m licenses sold] actually means. It doesn't mean that there are that many devices out the door."

According to the analytics firm, copies of Windows 8 sold to OEMs make up the bulk of the 100 million licenses sold. However, the 100 million mark excludes copies of Windows 8 sold to the enterprises as part of the volume licensing agreement.

Moorhead estimates that only 59 million Windows 8 devices are actually in use. That means that a little less than half of all of those 100 million licenses sold that Microsoft likes to brag about our sitting in a warehouse somewhere unused.

"How many Windows 8 PCs have sold and are being used?" asked Moorhead, arguing that that was a more accurate representation. "No specific news from Microsoft [on that] is bad news, because if it was good news, Microsoft would be shouting it at the top of their lungs. Instead, Microsoft is thinking they don't want to be persecuted over the number, so they're not saying anything."

Since Microsoft won’t offer specific numbers on how many computers have been activated running Windows 8, the estimate was based on data collected from analytics firm Net Applications.
 
Windows 8 currently accounts for 4.2% of all Windows PCs according to data collected by the company. The entire Windows ecosystem consists of 1.4 billion machines meaning that 58.6 million are likely to use Windows 8.

Source: Computerworld



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

RE: "Wild inaccuracies. . ."
By Mint on 5/13/2013 7:07:24 PM , Rating: 2
Yeah, if you're doing truly unbiased random samples.

There's no way for statcounter to ensure that.


RE: "Wild inaccuracies. . ."
By Schmide on 5/13/2013 8:02:43 PM , Rating: 2
Every sample has a bias Dude, that's the nature of the beast.

The issue was how many samples do you need to quantify the confidence factor. The above holds true for even the most biased sample.


RE: "Wild inaccuracies. . ."
By Mint on 5/14/2013 11:22:55 AM , Rating: 2
quote:
The issue was how many samples do you need to quantify the confidence factor. The above holds true for even the most biased sample.
Of course it doesn't. Biased sampling makes that calculated confidence interval completely meaningless. Go look at the assumptions in the derivation of the formulae you used.

Looking at all the hits to a particular set of websites is clustered sampling. On top of that, how you chose those websites is prone to bias as well.


RE: "Wild inaccuracies. . ."
By Schmide on 5/14/2013 4:22:06 PM , Rating: 2
I really don't understand your adversity? You seem to poo poo the data regardless of which type of opposition you choose, lack of samples vs biased data.

quote:
Biased sampling makes that calculated confidence interval completely meaningless.

Actually they are completely independent of each other!!!

Regardless of all this there isn't some magical error term that is going to sway these numbers in any dramatic fashion.


"You can bet that Sony built a long-term business plan about being successful in Japan and that business plan is crumbling." -- Peter Moore, 24 hours before his Microsoft resignation














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki