Stochastic error and biases remain in blind wine ratings

Abstract

AbstractAnalyses and aggregations of the ratings that wine critics and judges assign to wines are made difficult by stochastic error and biases that remain even when wines are assessed blind to price, label, capsule, and closure. Stochastic error is due to the partially random nature of ratings. Cognitive and omitted-variable biases are due to anchoring, expectation, serial position, commercial, and other factors. Differences in decanting, filtering, aeration, and temperature can also affect ratings.