As you are no doubt aware, the two ranking groundhogs, Punxsutawney Phil in Pennsylvania and Wiarton Willie here in Ontario, have decreed that we will have an early spring. Second-rater Shubenacadie Sam in Nova Scotia disagrees, but who's ever heard of him (or her)?
Yesterday evening on our local "superstation" CHCH, the newsreader was ribbing the weather guy over the fact that a majority of people trust the rodents more than they trust meteorologists, even though the success rate for Phil and Willie's annual prognostications is only about 37%.
"Thirty-seven percent?" sneered the meteorologist. "That's hardly better than random!"
Well actually, weather guru, it's quite a lot worse than random. The outcome here is purely binary -- either the ugly critter* is right or he's wrong. Over a long enough run (and Phil has been in business down at Gobbler's Knob for well over a century) a random outcome would produce a success rate of very close to 50%.
Scientific weather forecasting is largely about precedents and probabilities, so the meteorologist's ignorance is a bit surprising -- but it may help to explain why most people would rather trust the groundhogs.
* The groundhog, that is, rather than the CHCH weather guy.
No comments:
Post a Comment