What to Read Next
More than you might think.
In particular, people are gaining new insights about each of the three subjects through statistical models that analyze data to make predictions.
That was the message conveyed at a recent presentation by Dimitris Bertsimas, an MIT Sloan School of Management professor who teaches a course called “The Analytical Edge.”
Bertsimas, who spoke last week as part of a colloquium at the MIT Sloan School, described equations that help predict wine quality and baseball wins. For example, wine quality can be predicted through an equation that factors in winter rainfall, average growing season temperature and harvest rainfall. Bertsimas also explained that companies can analyze data from their employees’ healthcare transactions to better predict healthcare costs.
Of course, even good statistical models can’t anticipate everything. Bertsimas explained that last year, his baseball model predicted that the Boston Red Sox baseball team would win 90 games in the regular season — and they in fact won 89. This year, he predicted the Red Sox would win 101 games in the regular 162-game baseball season. But, one-quarter of the way through the season, the Red Sox had won only half their games so far.
“They are behind [the predicted record], I admit,” Bertsimas said.