[hoagsobject]

As someone who decided to get BSc degree in Math & Stats after 5 years in industry this thread makes me questioning my decision.

Everyone says you don't need degree, but in my experience if you want to do advanced stuff (machine learning, cryptography, algorithms) you probably need one, unless you are extremely talented and can pick up everything on your own.

You probably don't need degree if you want to work as SWE in mid-tier companies, once you have experience there are no problems with finding jobs.

[lordnacho]

> unless you are extremely talented and can pick up everything on your own

The question is more whether anyone will look at your CV if you didn't do these advanced topics at uni. Most things people can and do learn on their own. If you look at how uni works, what did you do? You went there and found out what you needed to learn, and then spent a bunch of time studying it on your own. There was never enough time in a class to learn things just by watching.

[hoagsobject]

It depends on the subject. You can get feedback from tutors/professors, and evaluation of your work. If you study proofs, you can't verify by yourself if they are good/correct.

Apart from that, if you go in a traditional uni (which is not my case, I do online degree at Open University of UK), you also have great networking opportunities.

Not saying about other subjects, like medicine, physics, biology where you need access to a lab to learn something.

[silent_cal]

No, I think you're right. It's easy to think you've understood these concepts when you really haven't. Sometimes formal training is really necessary I think, or at least is the fastest way to true comprehension. Especially with rigorous subjects like linear algebra and calculus.