I don't know too many people who would say analytics won Colorado the series, but it almost certainly helped them prepare and adjust.
But the more important point is that you keep asking for 'evidence' and 'facts', which is good. That's what everyone wants these days, and is what the article is really saying. In that sense, the data points are just little 'bits' of information, factoids in mathematical form, that teams study to try to extract meaningful evidence. It's just a modern day form of detective work, with lots of noise, but also signals in all the information collected.
If you believe in evidence as an organization, then you need to commit to gathering statistical information, then having experts analyze it, then translating it back simply to the people that need it to make decisions.