Wine Rating Data

Before the semester went fully online due to COVID, I had a lot of fun tasting wines in HADM 4300: Introduction to Wines. As part of the class, we tasted about six wines each week and recorded notes on them, including a 1-10 rating. Before leaving the class, I tasted and rated 32 wines from the US and France (they covered other regions later). After finding my notes from that semester, I thought it would be fun to do a bit of data analysis.

The big questions: White or red? France or USA?

I tended to prefer reds to whites and (ever so slightly) French wines to US wines. However, with such a small dataset, the differences are not statistically significant (t-test p-values of 0.12 and 0.97 for red/white and France/USA).

A closer look: Regions and varieties

Unsurprisingly, I liked Bordeaux reds (Cabernet Sauvignon over Merlot) and Loire whites. Perhaps somewhat more surprising is New York topping the average rating list, thanks to a really great Long Island Cabernet Franc and a pretty good Finger Lakes Riesling. Je suis vraiment désolé, Alsace, but I really didn’t like the funkiness of the Gewürztraminer and Muscat we tried. My favorite wine was a Loire Chenin Blanc, the only 9.5 I gave.

Other factors: Price and vintage

I don’t seem to have any particular preference for year or be influenced much by price (I gave all ratings price-blind). The Pearson correlation between price and rating is 0.19 (p=0.29). However, the two most expensive wines we tasted were both fantastic: Château Langoa Barton and Château Pédesclaux.

Conclusions

• Try more Chenin Blanc
• When in doubt, go for a Bordeaux Cabernet Sauvignon
• Need more data!

Raw data

In decending rating order: