Complexity increases forecast error

Complex models are more popular than simple ones among researchers, forecasters, and clients. The popularity of complexity may be due to incentives.

Less accurate

Kesten Green and Scott Armstrong propose that simplicity in forecasting requires that the method, the representation of cumulative knowledge, the relationships in models, and the relationships among models, forecasts, and decisions are all sufficiently uncomplicated as to be easily understood by decision makers.

Their review of studies comparing simple and complex methods found 94 comparisons in 29 papers. Complexity beyond the sophisticatedly simple failed to improve accuracy in all of the studies and increased forecast error by an average of 32 percent in 21 studies with quantitative comparisons.

Incentive to be complex

“Yet complexity remains popular among researchers, forecasters, and clients. Experimental studies suggest that the popularity is due to incentives”, according to Green and Armstrong.

They say researchers are rewarded for publishing in highly ranked journals, which favor complexity. Also, forecasters are helped to provide forecasts that support decision makers’ plans by complex methods and forecasters’ clients are reassured by incomprehensibility.

Read more…