In this video, get introduced to the concept of forecasting. Using Hurricane Matthew allows for an introduction to forecasting concepts with a medium most are familiar with.
- In the fall of 2016, Hurricane Matthew was making headlines around the country. The first Category Five Atlantic hurricane in a decade. State of emergencies were declared in Florida, Georgia, and the Carolinas. Throughout the storm's lifespan, meteorologists were constantly updating governments and the general public about the hurricane's most likely path. This process is called forecasting and is a form of predictive analytics. Forecasting is used by weathermen, pollsters, businesses and many more who wish to better understand how future events might play out.
The storm's path was often presented with a graphic that mapped out where the storm originated, where it is currently, and where it could possibly be going forward. At one point, forecasters had people as for north as New Jersey worried the storm would reek havoc. Why? A hurricane's forecast included New Jersey in its cone of uncertainty. Here, referring to the range of locations along the East Coast to meteorologists could envision Matthew hitting. Why were forecasters uncertain? Because there were so many variables involved.
The effects of other weather fronts, how the speed of the hurricane changed, et cetera. Because these factors could impact the hurricane at any, or all moments, the cone was widest the farthest from where the hurricane was. In other words, forecasters were comfortable making a strong and detailed prediction about the path of the hurricane through Florida when it was 100 miles from the coast, but they could only offer their best guess about where the hurricane will go after that. Because the farther in the future you try to predict, the worse you do. Sure enough, a day later, New Jersey was removed from the cone.
In other words, the chance of New Jersey being hit had been removed from consideration as a likely path for the hurricane based on the updated situation. Now at no point during this process did forecasters declare that New Jersey was certainly going to be hit. This is an inherent truth of all forecasting. It is usually most successful at reducing the number of likely outcomes. Meteorologists could rule out New Jersey as the hurricane moved up the Eastern U.S. Seaboard. And, providing the likelihood of different outcomes.
In this case, the high probability that Florida's eastern coast would be hit. In business, forecasts should be thought of in a similar fashion. A forecast is not going to determine the future. It will, if done correctly, reduce the number of potential outcomes into a cone of uncertainty. It gets bigger the farther into the future you attempt to go. Let's explore this idea a bit further with a case study.
- Qualitative vs. quantitative data
- Data analytics success stories
- Making predictions
- Asking the right questions
- Collecting data
- Understanding averages
- Sampling: pros and cons
- Cause and effect