3 ms·
You're missing the point. What you're really trying to do is statistically correct for "defaultness"; i.e. remove the users who would actually prefer browser X
by pcwalton 14y ago
You're missing the point.
What you're really trying to do is statistically correct for "defaultness"; i.e. remove the users who would actually prefer browser X but use browser Y because it's the default. Your methodology isn't a sound way of doing that, because it also excludes users who use the default browser on their operating system because they prefer it to all others. To be sound, your methodology would need a way to not count the former users while leaving the latter users intact.