Summary: Netcraft changes what it measures, which works pretty well for Microsoft; Net Applications carries on deceiving

THREE weeks ago we wrote about the mysteries of Netcraft, which is claimed to be bent in Microsoft's favour. A few days ago the following got published:

Netcraft: Microsoft Hosts Most Active Sites

Web analytics firm Netcraft (www.netcraft.com) is expanding on its web server survey by offering a dataset of the most active sites, giving a truly global analysis of web hosts.

The firm explains that each website that is found on its monthly web server survey can be attributed to a hosting location using reverse DNS and IP address delegation information.

Could Microsoft be compensating for this change in the element that's measured so as to glorify Microsoft's pathetic true numbers (10-20% market share, according to some sources)?

“They are linking to a cryptic survey that is indirectly funded by Microsoft and Apple.”The black art of statistics is associated with mastery of how to change what's displayed or measured such that it places the preferred data/company as leader and then presents 'quick results' (usually graphs) to unsuspecting viewers that are lazy to validate or to ask what it is that they are actually seeing, let alone how it was derived, bar caveats?

A few days ago we found Web sites which are pretending that GNU/Linux has only just surpassed 1% of market share on the desktop. They are all linking to a cryptic survey that is indirectly funded by Microsoft and Apple. For information about Net Applications, review past writings such as:

• Microsoft and Apple Pay Net Applications (Hitslink)
• Net Applications: the Big Lie, Boosted by IDC|IDG et al
• Summary: Lies, Damn Lies and Net Applications (Fake 'Statistics')

Regarding the latest from Net Applications, Bernard Swiss writes:

IIRC, Google Zeitgeist in 2003 reported 2.9%

and I trust that figure for client OS share much better than anything put out by NetApplications.

It's too bad that Google decided to stop reporting client OS share; I never understood why they felt it necessary to do so, but I suspect there was more to it than the official statement covered.

We have already shared our own perspective on desktop market share in [1, 2, 3, 4]. ⬆

"We don't have a monopoly. We have market share. There's a difference."

--Steve Ballmer, Microsoft CEO