Static analysis

From Wikipedia, the free encyclopedia

This article may require cleanup to meet Wikipedia's quality standards.
Please improve this article if you can. (May 2007)

This article is about the statistical technique. For the programming technique, see static code analysis.

Static analysis, static projection, and static scoring are pejorative terms for statistical analyses which take existing trends and project them into the future simplistically, or beyond what is possible to predict in any fashion, producing often wildly unrealistic results.

Its opposite, dynamic analysis or dynamic scoring, attempts to take into accord how variables may change or interact. One common area for the use of these terms is budget policy in the United States, although it also occurs in many other statistical disputes.

A famous example of static analysis comes from overpopulation theory. Starting with Thomas Malthus at the end of the 18th century, various "experts" have taken some short-term population growth trend and projected it for years into the future, resulting in the prediction that there would be disastrous overpopulation within a generation or two. Malthus himself essentially claimed that British society would collapse under the weight of overpopulation by 1850, while in the sixties the book The Population Bomb made similar doomsday predictions for the US by the 1980s.

Similarly, some scientists took a short-term trend of temperature declines in the 1930s and proclaimed that the world would be in an ice age by 1978. As with the overpopulation theories, the projection ended up less accurate than a roll of the dice because it didn't take into account how things interact, nor how a short-term spike was being treated like a long-term trend.

As previously mentioned, in budget projection debates, predictions that assume no significant changes in behavior in response to changes in incentives are often condemned as being static projection. The result of these scoring practices is that it tends to discount any policy change that would increase economic growth or enhance efficiency in government programs. By contrast, dynamic scoring refers to projections based on assumptions about the efficiency effects of policies such as tax cuts.

Typically, static analysis works for very simple systems: for example, how fast snow is accumulating in what is thought to be the mid-point of a blizzard. But even then it must be tempered with rationality -- guess how much longer the storm might actually last, rather than assuming that snow will fall continually for the next sixty years, and project the average of the storm so far, rather than plotting the curve of its growth as if that will continue to increase for the second half.

However, when applied to any more complex system static analysis tends to be worse at making predictions than the flip of a coin.

A possible instance of static analysis is the notion of technological singularity, whereby some factor of knowledge growth, such as computer intelligence, is projected into the future, resulting in exponentially rising curves that suggest that everything will be known by a relatively early date.

A satire on this idea has been presented using the development of safety razors: After their invention, all safety razors were single-bladed for 70 years. Then the first double-bladed razor was introduced. It only required 15 years for a third blade to be added, and then one year for the fourth and fifth. Fitting these five data points to a hyperbolic curve produced the prediction that within nine years of the calculation -- by the year 2015 -- safety razors would have an infinite number of blades.^[1]