Model Building as an Educational Hobby.

Abstract

The traditional hobby of building models such as ships and airplanes requires meticulous assembly of small-scale pieces in sequence. Many models designed to commemorate historic events provide education during the assembly, and those that are completed often become objects of display on a shelf. Model building can also become a career, using past events as a basis from which to predict future events of high commercial interest. Fatigue analysis of components in engineering, for example, has reached a high state of accuracy with narrow error margins, because the conditions of the testing can be tightly controlled to approximate the conditions of future use. However, using the past to predict a future with different conditions is flawed by wide uncertainty, as reviewed by Taleb in the Black Swan.1 Errors have been notorious for predictions from the economist, defined as an expert who will know tomorrow why the things he predicted yesterday did not happen today.2 Financial experts have been demonstrated to consistently underestimate the contribution of unknown factors to the accuracy of their predictions, which although often no better than amateur predictions are reliably accompanied by greater confidence, which increases with the number of variables modeled.1 See Articles by Upshaw et al and Lagu et al High levels of randomness apply also to both the individual human behavior and biological events related to heart failure (HF), creating even greater uncertainty at their interfaces, where outcomes of hospitalization and death occur. Two articles in the current issue of Circulation: Heart Failure provide insight into the construction and use of models for predicting HF events. Both emphasize the primacy of simple clinical factors available to the clinician. Lagu et al3 have compared models for death during hospitalization for heart failure, derived from data of various eras and tested in …