The chart there is somewhat misleading. It says "Usage share of desktop browsers" but NetApplications measures users, not usage. That by itself could explain m

The chart there is somewhat misleading. It says "Usage share of desktop browsers" but NetApplications measures users, not usage. That by itself could explain most of the variation in the chart, and lead one to conclude that IE has lots of users, but some of them do little browsing (imagine a grandparent that checks email or their stocks once or twice a day) while Chrome has users that browse a lot. Which is more important? Depends what question you are asking.