The trick for me is to learn a good foundation in statistics but to know that you don't need to learn everything.
A friend who founded a startup that makes heavy uses of statistics likes to trawl academic papers for algorithms that help his business.
Think of research papers a bit like a well documented private object/library -- you know what data it accepts, you know what it returns, but you don't need to know how it works.
Just make sure your code reflects exactly the formulae/model documented in the paper and you're good.
A friend who founded a startup that makes heavy uses of statistics likes to trawl academic papers for algorithms that help his business.
Think of research papers a bit like a well documented private object/library -- you know what data it accepts, you know what it returns, but you don't need to know how it works.
Just make sure your code reflects exactly the formulae/model documented in the paper and you're good.