Abstract
The author revisits the heralded 1976 Paris blind tasting of US Cabernets and French Bordeaux. The conclusion reported in 1976, 1999, 2001 and 2003 was that the Americans won the competition outright. This conclusion derives from a non-parametric analysis of the data that converts wine rating scores to ranks, thereby producing invalid results. The author discusses: the biostatistical importance of the reanalysis; the practical impact of the earlier conclusions upon the economics of the California wine industry; and the heuristic importance of the burgeoning body of data analytic strategies for further elucidating the scientific understanding of judges' evaluations of wines.