Now that the Philadelphia Eagles have won their first Super Bowl title, only one team with an all-time winning record has never won the Big Game. Coincidentally, that team is the Minnesota Vikings, in whose stadium Super Bowl LII was played Feb. 4. These factoids are just a small sampling of the obscure, yet interesting bits of information that people dig up out of the vast history of sports statistics.

John Stephens

Such bits of trivia are a frequent part of 24-hour-a-day coverage of athletic events on cable and across online platforms. Who doesn’t want to know that the guy just traded to another team is only the fifth person in league history to accomplish X after never having done Y, at least in the modern era?

OK, there might have been a couple of others who did it before they started keeping track of Xs as an official statistic back in 1954. These little tidbits come to light because someone was curious, and used analytical tools and technology to see if anything like that has ever happened before.

Similar methods are used in the insurance and health care industry to look for trends and patterns that can help manage risk and improve outcomes. It’s called “data mining” and it’s a way to gain value out of the enormous quantity of information that is collected in processing insurance and handling claims.

When we can locate insights against the background noise of billions of electronic transactions, we then have the potential to see if costs can be saved, performance improved, and possibly, catastrophe avoided.

Unlike the sports analytics used by coaches and general managers to make personnel moves or strategic game decisions, data mining is performed on a database that does not contain personally identifiable information about specific individuals. It does not compromise the privacy of anyone’s health or financial information.

Instead, the method associates pertinent data that may assist in finding root causes of accidents, determining the relative benefits of therapeutic drug treatments or understanding the circumstances that may lead a claimant to seek legal representation.

When enough insights are collected on a given issue, we can often develop predictive modeling to assess whether improved outcomes can be achieved in the future. The modeling can be used to evaluate effectiveness of case management in various illnesses or injuries, build incentives into benefit plans to utilize higher-performing providers, or measure procedures to reduce human error or mechanical failure. The ultimate objective is to help create safer, healthier workplaces and contain the costs of risk.

We are curious creatures to spend the time and effort on obscure knowledge about the events around us. New information will sometimes come our way and we’ll say, “Who cares?” Other times we’ll think, “How interesting,” but then we can’t remember when we have an opportunity to share it with someone else.

But we keep seeking, because for some of the obscure knowledge we find, there is someone out there who will say, “I’m glad someone was curious about that.”

John Stephens is senior vice president of Keenan and Associates' Public Agency Property & Casualty Practice in Torrance, California. This entry is republished from Keenan's blog.