...As Stephen Wright once said:
"You use statistics like a drunk uses a light post...for support rather than illumination."
Of course teams that hit 50% of their shots will likely win, there has to be a more telling stat than that, or at least one that is more of a stretch. At what point when doing this kind of numbers crunching do you realize your whole study is a 'duh-huh' and do you stop and decide to go in another direction?