The line about 'correlation' and 'causation' being two seperate entities is very true. Just because something shows a correlation does not mean that a direct cause has set the trend and that it will always be that way. It can just as easily be a correlation of coincidence.
However statistics, if analysed properly, are hugely important; especially in sport. Players are sold and bought and titles won and lost on the strength of properly analysed statistics. But analysing staistics properly isn't always, or is rarely, done right. Every single variable or reason and anomaly has to be looked at and analysed too, not just some but all. If you can't get all you can't analyse the statistics fully. Everything has to be taken into consideration.