Although I touched on the sport in last week’s blog, since this week is the real start of the Major League Baseball season—despite the fact two American teams played in Japan last week (with a Japanese tech company’s logo on the batting helmets)—it’s time for my annual observation about analyzing data in different ways.
Since last season, the predictive analytics book Moneyball was turned into a movie and both Brad Pitt and Jonah Hill were nominated for Oscars for their roles, which should have excited data geeks. After all, when you can get Pitt to portray someone who believes in statistical analysis of data you’ve won half the battle, right? (He could do for actuaries what Julia Roberts did for paralegals.)
The benefit insurers have gained over baseball teams in the use of advanced statistical data is that pretty much all insurance companies believe there is value to be gained from the data. There are plenty of baseball executives and managers that only believe what they see on the field—even if they don’t always know what they are looking at.
As user-based insurance expands across the personal lines insurers—my agent tells me Nationwide will be coming out with a new personal auto telematics product this summer—it will be interesting to see how they interpret the driving data they receive.
I spoke with a friend about telematics a couple of weeks ago at the IASA Boot Camp and she said she was not looking forward to widespread use of telematics because she admits that she drives too fast.
The safe assumption is that people who drive fast are bad drivers. But what if, over time, the statistical data insurers accumulate through telematics leads us to findings that we didn’t expect.
Just as baseball managers continue to waste outs by having good batters lay down a sacrifice bunt with a fast runner on first base, will insurance companies continue to believe that those who have too many speeding tickets are bad drivers and that they don’t need telematics to prove such a belief is true—or maybe even false.
We shouldn’t be afraid of collecting more data; we should be afraid if we assume that new data always verifies old beliefs.
So let that batter swing away with a runner on first base and let telematics tell us what the real factors are behind poor driving that leads to accidents and, inevitably, claims. We might all be surprised what the predictive analytics of that data tells us.