Texas Weather and Data … Get Your Umbrella Ready?

Although I live in New York now, much of my life, particularly my secondary school and college years, I lived in Texas. It’s the state I identify with and where my parents, brothers and sister live. This year, when I would speak with my family, what inevitably came up was the bad weather. I can’t help but laugh since I get about 5 weeks of great weather in NYC each year, otherwise, it’s either snowing, raining, melting heat or a thick cloud of humidity covering the sidewalks you have to walk on. Still, in the winter, all the talk was about their snow. Now, it’s all about the rain.

471938159_ca4f2c627d_o

Flooding seen throughout Texas not that predictable. (Photo by Rich Anderson)

It’s with good reason, though. This spring, the Dallas-Ft. Worth area had its second wettest stretch of months on record. In May, the state surpassed the previous record for rainfall by 32%. Heading into spring, nearly all of Texas was either dry, nearing drought conditions or in drought conditions. Now, barely any of the state needs more rain.

Flooding, however, has accompanied this historic rainfall. Videos from Austin to Houston to Dallas show river-rapid-like conditions in areas where you might have previously mistaken the water-flow for a creek. This has had tragic side-effects, including people losing their homes and even a number of deaths. And now another threat of flooding comes as Tropical Storm Bill makes landfall. Hopefully, everyone can stay safe.

But the reason for the weather is because an El Niño formation, leading to the massive water dump. In December though, there was only a 58% to 65% chance that such a formation would form. So just better than a coin flip. Certainly not a warning that would convince people to build an ark.

Of course, the old adage goes ‘you can’t predict the weather,’ but with the growth of data and analytics, that’s exactly what people expect. It’s the same in organizations, where the use of data has increased the desire to pull any and every piece of information a company can on customers, clients, prospects, and employees. It’s an attempt to predict the future. Or sway it. But it’s also an effort to reduce risk. After all, if you know everything about an employee, there’s no way they can disappoint, right? If you understand all of your clients needs and efforts, even those s/he doesn’t know about, then there’s no way they will ever want to leave. If you provide prospects with something they don’t even realize they want, there’s no way they will pass. If only that’s the way things worked.

That’s not at all what happens because with all this data, what we get are statistics. It gives us a range to work within, saying X is unlikely and Y will almost always happen. The problem with this mindset, though, is that humans, in the right circumstances, can fall within the small percentage of X or not live up to the high percentage of Y.

You see this in sports all the time. For instance, the chance that Golden State Warriors would win the NBA Championship Finals was 76% heading into Game 1 with the Cleveland Cavaliers. Nearly every prognosticator had Golden State in a landslide, except for those betting on LeBron James. The prognosticator’s were right. But in the same NBA, teams systematically lose on purpose in hopes of creating a better chance for a high draft pick. The Philadelphia 76ers and the New York Knicks planned for top one or two picks. Instead they got numbers three and four, outside the zone where they can grab a potentially franchise-changing center. Smart play or a fool’s hope? The answer might depend on how you feel about stats.

In the end, these numbers often predict only when they’re right. Otherwise, the failure was accounted for in the percentages.

That’s one of the problems with data. It either gives us unfounded confidence or lulls us into not preparing for the unexpected. We see that it’s a 90% chance that Google will rise, and we want to buy. But that 10% is still real. And in between that 90% chance of a win and 10% chance of a loss is where real life happens, moving those statistics far beyond what we can predict at this moment.

Sure, one day data may tell us exactly what will happen and how. I can’t imagine the boredom that will bring, since without the risk, we don’t see the reward. But I guess we’ll have plenty of time to prepare for rain.

 

Posted in Data Tagged with: , , , , , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *

*