Sure, but Ruby could've been a player in data science as well. But for some reason Python won handily in that space. Was it all down to numpy, scipy, etc?

Whenever this subject comes up, people seem to forget how poorly Ruby plays with Windows. WSL helps a lot and gives you a practical experience, but it'll never be a first-class citizen on Windows like Python is.

Maybe, but most data science jobs I see (and have been around) seem to be using Linux... with a few on MacOS. Lack of Windows support doesn't seem like a problem in the realm of data science.

With 80% of the desktop market, most IT shops work on Windows desktops before submitting their work into a Linux cluster.

I think it's because ruby got pigeonholed.

Working with ruby became synonymous with working on rails to the detriment of the rest of the ecosystem.

Additionally, I think there are a lot of non-programmers, non-engineers working in data analysis or data "science" and Python syntax is much more approachable for a beginner.

Lots of starter programming courses whether university or online are in Python, etc. Again if you're not actually and engineer or developer of sorts, you're not likely to spend much time and effort moving on from the language you already know that everyone else in your field using, etc.

Yes, I think numpy, scipy, and pandas were key.

The answer to that is yes. Python was already used a lot in science and when Data Science started to blow up it already had numpy, scipy, scikit-learn and pandas. Also those libraries are really fast because they're wrapping highly optimized Cython, c, c++, and Fortran. I don't think Ruby had anything like that at the time.

Any language with FFI can have similar bindings to the same libraries used by Python, see Java, .NET (C#/F#), Julia, Swift,..... yet it doesn't seem to happen for Ruby.

As someone who makes a record high salary doing Rails development, I beg to differ.