OK, there's a third option: Figure out WHY you got the goofy result, but that essentially the same thing as option one, assume you then tweak the formula. After all: If you continuously tweak the formula until its gives you an answer that you already "know" is right... well, that's no better than just picking based on your own opinion and personal feelings, is it?

The reason I call this the "Candy Maldonado rule" is because of a conversation I was having years ago about this very thing. My point was, that is you have this formula, that you really like, but it tells you that Candy Maldonado is the greatest hitter that ever lived, it's a garbage formula. How do I know? Because Candy Maldonado can't even shine the shoes of the 100th greatest hitter of all time. How do I know THAT?

I. JUST.

**KNOW**.

And, with all due respect to Mister Maldonado, his was just the first name that came to mind. Believe it or not, it's pretty hard to think of someone who just plain SUCKS right off the top of your head. And the first name that comes to mind when you're trying to come up with one, is usually someone who was actually pretty good at one time. And for some reason "Candy Maldonado" just popped in to my head. And he

*was*

__good__. Not very good, and certainly not great, but...

__good__.

And I thought it made a really memorable name for the rule and it stuck. So for all the statistician's efforts to take subjective opinions out of the analysis of the game, the results of these formulae can not just taken

*blindly*. Each formula must still judged against

*common sense*. And that basically means: It better pretty much tell us what we already "know," with no more than a few surprises.

IOW: It has to pass the "Candy Maldonado rule" and not give a truly absurd result like "Candy Maldonado is the greatest hitter of all-time." Ruth, Bonds, Williams, Cobb, Hornsby, Mays, DiMaggio, Aaron or Mantle? Fine. Maybe you have a good formula. But if Candy Maldonado breaks the top ten:

**You don't**.

And no, I can think of no formula that would reach this conclusion. It's just what I call the

*common sense test*.

BACK