Tasting order and wine scores

It is well known that your perception of a wine is affected by what was tasted immediately beforehand. This is similar to how other senses behave. If, for example, you sit in the dark for several minutes and then expose yourself to normal daylight, it seems to be exceptionally bright until your eyes adjust. With wine, if the previous wine was flabby, your current wine will tend to taste sharp; if it is was very dry, then sweet, etc.


However, in addition to that effect, the order in which you taste wines can help determine how much you like them. I have already written about Antonia Mantonakis’ research, where she presented sequences of wines to consumers, and asked which wine they liked best. In actual fact the same wine was offered in each glass, but the tasters still expressed a preference. For shorter sequences (2 or 3 glasses) they tended to prefer the first wine tasted, while for longer sequences (say 5 glasses) they preferred the last one. If you are thinking “that just goes to show how little the average consumer knows about wine”, you should be chastened by the fact that those with better wine knowledge were even more prone to this bias. But how much practical significance does it have? The consumers were after all asked to distinguish between identical wines, and they may have thought any differences were very small.

Later work[1] by French researchers throws more light on this order effect. Here the experiments were performed on wine professionals  tasting in competitions that awarded medals to Beaujolais Nouveau wines. Here, in each of two competitions of around 400 wines, approximately 100 tasters scored wines on the 100 point scale, each taster being given two flights of between 10 and 12 wines. But into each tasting sequence of wines registered in the competition, the experimenters added the same wine (an unregistered one) into the first and penultimate positions. For each competition, the average score for the penultimate wine in the sequence was greater than that for the first wine: 82.99 vs 79.78 for one competition, and 83.93 vs 81.78 for the other. Those differences are not only statistically significant, but also significant from a practical point of view. To get some feel for the practical significance, note that the difference between a silver and gold medal winning scores is 6 points. Also, as the required score to achieve a silver medal was 81, in one of the competitions the position of the wine would have decided whether the wine received a bronze or silver medal.

If wine scores and medals are important to as a consumer, then you should probably be concerned about the sequencing used in tasting flights. On the other hand, if you do not pay much attention to such things, you can now add order effects to the list of reasons that justify your position.

[1] Carole Honoré-Chedozeau, Jordi Ballester, Bertrand Chatelet and Valérie Lempereur, “Wine competition: from between-juries consistency to sensory perception of consumers”, BIO Web of Conferences 5, 03009 (2015)

Order effects, and linguistic fluency

I thought that subject line would get your attention 🙂

In you have an hour or so to spare, take a look at this lecture by Antonia Mantonakis entitled “Does a wine’s name influence consumer taste perception?”  It also covers other influences on taste perception that have nothing to do with the wine itself.

So, to my list of factors that explain why wine tastes the way it does, it seems we can add the following two:

  1. Order effects
    I mentioned in my earlier list that the previous wine tasted affects how the current one is perceived, but Antonia found more general order effects.  In short sequences of  up to 3 wines, consumers show a preference for the first one.    But for longer sequences, up to 5 wines, the last one is preferred.   This research is briefly referred to in the video, but I could not find a publicly accessible research article to link to.
  2. Linguistic fluency
    Two winery names were made up, one of which was easier to pronounce than the other, but in other respects they were the same. In contrast to how linguistic fluency (ease of reading and pronunciation) affects perception of more basic consumer goods, it was found that a wine associated with the less fluent name tasted better.   More details in the video and this article.  I am sure I have heard it said many times that wines with simple names have an advantage in the market, but maybe that idea came about by the invalid extension of results from other product categories.

I’d like to emphasise that these results refer to reported perceptions.  There is admittedly the issue of whether people say what they really think about a wine, or say what they think they should be saying.  It might even vary from experiment to experiment, but MR scan evidence and a clever experimental design suggest that people at least sometimes say what they really think they perceive.

The final thing I would like to comment on is perhaps the most surprising, at least for those who fancy themselves as wine experts. The subjects in the experiments mentioned in the two points listed above were divided into two categories: those who knew more about wine, and those who knew less. Guess which group was more influenced by the order and linguistic fluency. It was the more knowledgable one!