Statistics and mathematical modeling have changed the way society functions.
Meteorologists use new methods of gathering information, especially from the atmosphere, to create a far grander database than any that human society has used for that purpose before. They feed these data points into supercomputers, and create complex models that involve a multitude of variables — often variables that are in a constant state of flux or that are difficult to get firm information on.
Edward Lorenz, who served as an army weather forecaster during World War II, and then studied and taught meteorology, coined the “butterfly effect” as shorthand for how tiny variations could affect weather models: “Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas?” He pioneered the field of chaos theory largely in an effort to better predict meteorological changes. Scientists continue to use long-reaching models to predict damage caused to the earth by global warming and human impact, while similar (albeit less complex) models are used all over the Earth for long-range forecasting.
The world economy, tracking the financial dealings of seven-billion plus through a dizzying array of legal jurisdictions, tax laws and the like, is tracked through a series of macroeconomic statistics; the ones produced at Berkeley use GDP, CPI, unemployment rate, corporate profits, change in business inventories, housing starts, interest rates, exports, and personal savings rate, among others. It’s on the basis of similar economic models that men like the Governor of the Bank of Canada or the Chairman of the Federal Reserve issue projections and make their decisions — decisions that affect millions.
Even the advertising industry has made extensive use of mathematical modeling. Taking statistics compiled by government and the private sector, they break the population into demographics, and target their ads to different sectors of the population. By doing this they hope to better market their products. In 2001, The Coca-Cola Company alone spent nearly $2 billion on advertising — money that was largely distributed on the basis of demographic information compiled by market research companies.
Even sports teams have turned more to mathematical models in recent years. The book Moneyball, based on the success of Oakland Athletics’ GM Billy Beane, is generally considered to be the work that brought statistical work into the mainstream in baseball. Theo Epstein, originally a PR man with the San Diego Padres, worked his way up the ladder and in 2002 was hired by the Boston Red Sox as General Manager. He was 28; the youngest GM in the history of Major League Baseball, and a man who’d never played the sport at even the high school level. Using sabremetric principles (math pushed by the Society for American Baseball Research), his team has won two World Series championships in his six years at the helm.
These are just a few examples; in virtually every scientific field, whether “hard” science or the social sciences, mathematical modeling is a primary tool for extrapolating all kinds of factors. It’s used in other sports, it’s used in scenarios that are for less predictable than what occurs in the carefully regulated confines of the arena, and often it’s used with data far less complete than we have available for free from the stats page and play-by-play reports at NHL.com.
In most of these other fields, suggesting that your powers of observation and gut feeling were a better method for prediction than mathematical models using a vast database would get you laughed out of a job.
In hockey, it’s the only commonly accepted practice — and I have only one question. Why?