A study entitled “Does Blind Tasting Work? Investigating the Impact of Training on Blind Tasting Accuracy and Wine Preference,” published in the Proceedings issues of this journal, analyzed the effects of training on blind wine tasting accuracy (Wang and Prešern, 2018). I point out two issues with that study and reanalyze their data. I find that the effects of training on accuracy are small, even without controlling for self-selection bias that may produce upwardly biased estimates. To the extent training works, it does not seem to work well and it may only work as a selection device. (JEL Classifications: C91, D83, L66)

We analyzed data from Oxford University Blind Tasting Society's 2018 training season to assess whether blind tasting training improves accuracy. Over time, guesses for grape variety increased in terms of accuracy as well as within-group agreement. Moreover, for grape variety, location, and vintage, the chances of the most common within-group guess being correct were significantly higher than the underlying frequency distribution. Finally, we observed a shift in preference towards older wines, with those with little initial experience gaining a preference for greater acidity and alcohol, and decreasing their preference for oak. Our results have important implications for growing wine markets with an increasingly educated consumer population. (JEL Classifications: C91, C92, D83, L15, L66).