Causal Analysis: The Buoy Pop

I spend a lot of my time looking at data. Sometimes I get the chance to look at trends and perform causal analysis. What is causal analysis? It is looking at data and figuring out when one event causes another. A good example is that is the old-time belief that the rooster crowing in the morning causes the sun to rise. Don't laugh, that was an actual belief. We all know that it is the rising sun that causes the rooster to crow and not the other way around.

A more complex example of causal analysis is in correlating wine drinking with good health. There was a study done that indicated people that drink a glass of wine a day are more healthy than those that don't. On the surface it would seem that everyone should drink a glass of wine every day. Unfortunately life isn't that simple. There are a number of other factors that need to be taken into account as well. The reality is that if you can afford to drink wine, you can also afford proper healthcare. It is not the glass of wine that makes someone healthy, it is affluence and the ability to get proper medical attention when needed. They never mentioned that in the study.

Yesterday I came across a very interesting article about the "Buoy Pop." There is a scientific buoy in the middle of the Pacific Ocean that can predict a snow storm in winter for Utah 2 weeks before it happens. This is important if you are a skier and want to know when the next powder storm is on its way. Meteorologists discount the buoy because the best they can do is predict storms 10 days out with only 20% accuracy. How can wave height indicate a storm 14 days away? This is where data science comes in and says there is some other phenomenon at work that meteorologists have not yet identified. Before continuing, the "Buoy Pop" is only 66% accurate, not 100%. Still that is significantly better than what professional weather forecasters can do.

In order to do proper causal analysis, I would need to gather all possible variables and apply a number of statistical algorithms to see what the actual cause of the snow storm is. All I know for now is that something that causes the storms to arrive also sends a nice wave under a specific buoy in the middle of the Pacific Ocean.

Ultimately the "Buoy Pop" is a nice statistical problem for me to think about but it doesn't effect my actions much. I purposely live in Utah near the ski resorts so that I can take advantage of powder storms as they arrive. For someone who has to travel to ski, I would definitely study when the buoy pops and plan a trip to my favorite Intermountain ski area.