So Here's My Idea:

A little under a week ago, I read a weather report that said Sunday's weather had a 100% chance of rain. 100%. The forecaster was so sure it was going to rain, he had the audacity to bust out the big 'hundo -- and then it didn't rain. Sure, there was a massive pineapple express queued up out in the Pacific, and sure, it was pointed straight at us. But somehow, the fates intervened, and decided to pummel SoCal with our rain instead, leaving Seattle with sunny skies and temperatures nearing the sixties. Thanks SoCal -- we owe ya one. The point is, even when the professional meteorologists predict something with 100% certainty, there's still enough chaos out there to flummox the entire model. (This is supposed to be a science, right? Shouldn't they have learned to avoid using 100% chance of anything by now?)

This fundamental unpredictability reminded me of something: the stock market. I've never noticed the connection, but the two systems seem quite similar: both sets of analysts are using a limited amount of data to make predictions about a system of staggering complexity. In both cases, little details can warp the whole model, horribly skewing the real outcome away from the predicted one. Here's my idea: why not let them swap jobs? Would anyone notice? I doubt it.

I also like the idea of making weather predictions based on things that have nothing to do with weather:

CHICAGO, Jan 15 - US wheat futures fell 3.4 percent on Friday, hitting their lowest level in more than two months, increasing the likelihood of a wet weekend for us here in Illinois, traders said.

Corn futures were down 2.5 percent, falling for the fifth straight day, indicating an almost certain cold front approaching from Canada. If the trend continues, we could see temperatures in the low teens as soon as late next week.

Meanwhile, turning to National weather, the Dow rallied unexpectedly yesterday, closing at 11,486 with a gain of 122 points, which implies a high pressure system is building over the Gulf of Mexico, reversing last week's predictions of a monsoon in Kentucky.