Explaining Probabilities

The National Weather Service has been toying with us this week, throwing out percentages and probabilities that the next flash flood will hit your back yard. Because of this, I have deemed it time to repeat a blog entry from last year that explains, for instance, exactly what a "40%" chance of rain is. Here we go:

Let's break down the numbers and what to really expect. Remember now, this is groundbreaking stuff. I could have kept this kind of breakthrough material under wraps. But I'm first and foremost a public servant. Even though I could sell these secrets for many millions, I choose to stay humble and not mercenary.

10% chance: Rarely used and seen only in forecasts put out on the NWS's website. It indicates a total lack of confidence in rain chances and is used when a few cumulus clouds will interrupt a clear sky. Furthermore, I contend it shall not ever rain with a 10% chance.

20% chance: This is a cover-your-bases percent used by the NWS that unnecessarily gives hope to the masses. It should be discarded and never used. It mainly is used in August when the NWS knows that out of the 7500 cumulus clouds, two are going to produce rain. So they string along the public much like a 2 AM infomercial.

30% chance: I consider a 30% chance really half that. The NWS nerds use it when some over-estimated factor has grabbed their imagination...usually in the summer as they bow to public pressure to DO SOMETHING!

40% chance: This is a good percentage. It rains 62% of the time when there's a 40% chance. There is no explanation for this other than NWS incompetence.

50% chance: Not so good. Part of the problem is with the fact that "5" is an odd number. Odd numbers are not good. September 11, 2001 is really a combination of three odd numbers...9,11, and 2001. Crazy Chrissy Columbus "discovered" America in 1492, an even number. I was born on 10-12-48, three even numbers. I've made my point. My studies show that with a 50% chance, actual chances are really 31.3%, an odd number.

60% chance: Prepare for a good, soaking rain. My data says it'll rain 82.24% of the time.

70% chance: The public gets all excited with such a high number. They cancel picnics and outdoor weddings. They storm the hardware stores (so to speak) for rain gauges. Surprisingly, this is one guess that the NWS scores on. Personally, I think it's just God taking care of people less smart than I.

80% chance: Bingo! This is the one you pray for. It's an even number. And my research shows it's a sure thing. Not an 80% thing, a SURE thing. Need proof? I went through microfilm libraries around the world and discovered (amazingly) that there was an 80% chance of rain for these well-known events:a. the Johnstown floodb. Noah's flood (an unprecedented 40 straight days and nights of 80% chances)c. Katrina

90% chance: This is a little creepy. It is rarely used and the great unwashed public (get it?) hardly knows what to make of it. They wonder, "If it's almost a certainty, what is it lacking?" Doubt clouds (get it?) their minds and gives way to unabashed hand-wringing (get it?). And with good reason. This odd number percentage has produced rain exactly 47.313% of the time since statistics have been kept. (Statistics have been kept since Joseph interpreted Pharaoh's dreams and Potiphar was given the first annual Troy Dungan award).

100% chance: Amazingly, this is the 2nd biggest fraud ever perpetrated on the American public. An Oak Ridge Boys tour is number one. This percentage is the kiss of death. It never rains when the NWS is suckered into this "sure thing". There are no sure things, except an 80% chance of rain, or Richard Sterban showing us his chest hair (singular).