It is that time of year again and I love it. March Madness is one of my favorite times of year. By now, many of you have probably filled in your brackets for the NCAA tournaments. In my family, we enjoy doing brackets for the men and women. Here’s a question for you. Would you be happy if 90% of your picks were accurate? Of course you would. That’s actually how accurate a 5-day weather forecast is today so enough with the cliche and misguided statements about weather forecast accuracy.
As a meteorologist, I am quite familiar with wisecracks about the accuracy of weather forecasts and ever-present statements like, “It must be nice to be in a profession in which you can be wrong 50% of the time and still get paid (insert eye roll).” As I have written several times, that statement usually tells me more about the person uttering it than my actual profession. The public generally struggles with concepts of probability, uncertainty, and statistics. Yet, many weather forecasts and messaging use them. For example, precipitation is often given in terms Probability of Precipitation (PoP), which according to the National Weather Service website, “Describes the probability that the forecast grid/point in question will receive at least 0.01″ of rain.” During hurricane season, a cone of “uncertainty” and other probabilistic graphics are used to characterize potential locations for storms or their associated hazards. As we saw with Hurricane Ian (2022), there is great confusion about what information these graphics are conveying. Many people erroneously assume that with the “cone,” for example, the storm is predicted to go down the center. Such misinterpretations can have dire consequences.
I have found that the aforementioned concepts, misuse of Weather App information, and the tendency by people to remember the smaller sample of missed forecasts (and forget the more numerous correct ones) leads to a perception that weather forecasts are bad. They are actually pretty good. The National Oceanic and Atmospheric Administration (NOAA) SciJinks website provides a pretty straightforward analysis: “A seven-day forecast can accurately predict the weather about 80 percent of the time and a five-day forecast can accurately predict the weather approximately 90 percent of the time. However, a 10-day—or longer—forecast is only right about half the time.” That’s pretty good folks and within 5-days the accuracy can often be better than 90% so resist the urge to form conclusions about weather forecasts based on the 5-10% misses or the rainfall that spoiled your cookout. Instead, it is important to ask yourself whether you understood the PoP that day or only looked at the “Smiley Sun/Cloud Face” on your App without properly synthesizing the rainfall scenario over the 24-hour period.
Now, let’s go back to your March Madness brackets. According to statisticians, the odds of picking a perfect bracket are 1 in 9.2 quintillion. You are more likely to be struck by lightning or one of those mega lotteries. Perfect brackets are unlikely. As a matter of fact, I would guess that the majority of us will not come close to getting 90% of our March Madness picks correct. When you look at people who try to forecast the future (sports analysis, stockbrokers, political pundits, doctors, and sports fans), I’d say my profession more than holds its own.
Happy March Madness.
Leave a Reply