Is PPC management much more than a crap-shoot?

I’ve been managing PPC campaigns for a few months now, and I’m still always learning new things. I’m still debating as to whether it would be more beneficial to go “all out” and adjust max cpc’s to keep ads ranking in the top three positions on only the “cream of the crop” keywords; or to target more mediocre keywords (yet more precisely targeted) that still gain healthy traffic. When it all comes down to it, I’m learning that affective PPC management is about 50% strategy and 50% things that are out of your control. Thus, tracking the effects of changes made to a campaign are key. And it’s also important that you establish a solid linear way of analyzing data, so you don’t end up comparing apples and oranges for insight.