Continuous multivariate probability distributions are highly relevant to many important decisions, such as how to best allocate resources across a portfolio of correlated financial or R&D assets; which of multiple oil fields to develop; and choice of aim point in sports, military, or space-exploration applications. Traditionally, however, many of us have avoided such distributions because they have been impractical to parameterize, validate, and/or simulate. Here we introduce new practical methods and apply them in a new application area: strategic decision-making in golf.
Because of their desirable properties, quantile-parameterized distributions (QPDs), which include the metalogs, are being used increasingly instead of traditional distributions (normal, student-t, beta, etc.) for practical applications of probability. Application areas include not only decision analysis but diverse fields ranging from environmental conservation to astrophysics.
The metalog distributions and other QPDs give rise to improved multivariate methods -- enabling such distributions to be more accurately specified, easier to elicit, easier to determine from data, easier to validate, and easier to simulate. We illustrate these features with an easy-to-understand practical application: in golf, choosing the best aimpoint can make the difference between winning and losing.
