I remember back when Visual Studio 2012 shipped with black & white toolbar icons, and main menu in ALL CAPS. This was all backed by studies that supposedly showed how most users prefer it that way, according to the lengthy blog post about it.

Let's just say the community's reaction was not in line with those studies...

Testing and metrics can be wrong like in this case. A ton of data can be too much where we find patterns to reenforce what we want.

I'm not saying that you're wrong, but what Microsoft was doing is closer to science than what HN does every time an article somewhat related to the Ribbon is posted.

I'd bet on Microsoft's approach over these rants, every day of the week.

Each person who voices an opinion is a real value and a real data point. What Microsoft is doing may look like a science but they are not asking the right questions so they get misleading results.

Reminds me of when facebook pushed autoplaying ads. A/B metrics when through the roof. Sales never saw so much engagement. Every slapping each other on the back. Ask a user they didn't understand what was happening, there was a fear that was introduced around scrolling your own feed.

Science is great but only attempts to answers the questions you give it.

I recall the first time I used the new windows 8 UI. It was absolutely horrifying. There were 2 different control panels with different settings sometimes overlapping. The traditional UI in which you could start apps by clicking on desktop shortcuts and manage traditional windows but certain apps including the start menu put you seemingly randomly in the weirdo new app experience where the window took up the whole screen. This was also true of clicking a url link in a different app.

It was bar none the worst desktop interface that anyone has ever released for a computer.

You have asserted without evidence that MS's UI is driven by science and you are assuming that Microsoft's research is in fact any good. Based on consistently mediocre results I have reason to doubt that. To support your position I suggest you actually cite some research instead of assuming it has been done competently.

I read those blog articles as they came out, and bought into the same kool aid as you did, but when time came to use it, it just didn't work as well as menus.

The contextual toolbar is one of the better ideas though.