Board Thread:Technical Updates/@comment-957747-20190910193755/@comment-5288784-20191015122341

I don't have exact ratios that are used as cutoff points, but your simplification would work, if we also assume that the total count (that is: all browsers, including the popular ones) stays the same.

Now let's assume 5% is the cutoff point, "Chrome" will represent one super popular browser that always appear in the stats; "Firefox" this time will represent a browser that has a small group of regularly returning, consistent users; and "Hipsters" are people using various other browsers.

What happened here? Even though the number of Firefox users and hipsters stayed exactly the same, a ton more Chrome users decided to visit the wiki. This alone caused the Firefox percentage to drop from 10% of total users to a mere 2.5%. Since it dropped below (our hypothetical) 5% cutoff point, it was then added to the hipsters group and displayed together as "Other". If more Firefox users appear or some users of different browsers stop visiting, Firefox will once again be displayed as normal.