On National Rosé Day (June 9th), I invited a few writers here in Houston to my house for what I believe to be was the largest blind tasting of American True Rosés in history. What is a “True Rosé”?
Well, there are essentially three ways to make a rosé wine. The first, which is rarely practiced outside of sparkling wine production, is simple blend of red wine and white wine. The second, which is widely practiced around the world, is called the “Saignée Method” where shortly after a red grape crush, a portion of the grape juice (after brief contact with the skins) is bled off (“saignée” means “bled” in French). This bled off wine is then vinified as if it were a white wine.
The third option is what I call a “True Rosé.” In this process, the grapes are raised, picked, and processed with the idea of making a rosé in mind. True Rosés are therefore not a byproduct of red wine production, they are intentionally or purposefully made. They are True Rosés.
Since this is my last post on the tasting that we staged at our humble abode this past June, I thought I would write a bit about my general perceptions, my pre-conceived notions, and end with a healthy dose of statistics (which will likely put most to sleep, but is firmly in my “geek zone”).
Simply put, there is no getting around the fact that 68 wines at one sitting is a sheet-ton of wines to try. No I know that there are a ton of “professional tasters” that are just getting warmed up after 60 plus wines, but I will reserve judgment on what I think about that practice (it’s stupid—so much for being reserved).
Tasting that many rosés in a short time, I learned a few things:
- There are a finite number of ways to describe the color “pink.” I tried to get creative, I tried to be introspective, I tried to be whacky. At the end of the day, though, there is pink, there is orange, red, and their combinations. Any more than that, you had the 256-count box of Crayola.
- Similarly, there are a finite number of “red-berry fruits” on this planet. I like to think I have tasted and smelled a majority of them, but that leaves me with six: strawberry, cherry, raspberry, blueberry (more “blue” than “red” but rock with me), blackberry (see previous parenthetical, swapping “black” for “blue”), and boysenberry (while I have never actually seen a boysenberry, I am a fan of Dannon Yogurt, so there you go).
- Acid is no joke. For a solid week after the tasting, my teeth hurt. They hurt to the point of contemplating a visit to my dentist. While he is a perfectly nice guy, he is also a Sadist who like to lecture me on how bad my oral hygiene is. So yeah, I avoid that dude.
- I thought I would be able to tell which wines were True Rosés, and which were saignées. For the most part, that was true, as I correctly identified more than half of the saignées. (Although, there were a mere handful of them so read into that what you will.)
- In a similar vein, I thought it would be fairly easy to identify whether a wine was a current release (i.e., a 2017) or an “older” wine (pre-2017). Well, here, that was much more difficult as I was right far less than half of the time.
I contemplated running a host of complex statistical analyses: ANOVA, non-parametric t-tests, even various regressions (linear, multiple, logistic, stepwise). In the end, I decided on running several correlations since they are the easiest to understand, and I no longer have access to sophisticated statistical software.
Basically, a correlation is used to determine whether two variables are related or connected (the word interdependent is also used). The test used results in a number between 0 and 1 where zero indicates no relationship and one defines a “perfect” correlation. The number can either be positive or negative, depending on the nature of the relationship. In a positive correlation, as one variable increases, so does the other (the longer you ride a bike, the more calories you burn), and a negative correlation implies an inverse relationship (the more you drink, the less coherent you become).
There are certain thresholds as well. A “strong” correlation is when the r (the variable used in correlation) is 0.7 or above. A moderate correlation is when the r is between 0.5 and 0.7, and a “weak” correlation is between 0.3 and 0.5.
Our stats class over, here are the correlations that I ran:
The relationship between age of the wine and relative quality (as perceived by me):
r = – 0.31
Retail cost and perceived quality:
r = – 0.004
Perceived quality and order in which the wine was tasted:
r = – 0.26
There is a common phrase among statistical types: “correlation does not imply causation.” In other words, just because two variables are correlated does not necessarily mean that one causes the other, but there does seem to be some sort of weak relationship between perceived quality and age of the wine (older rosés generally rated higher). An interesting finding for those who think that rosés should be consumed as young as possible (and lends some credence to my theory that True Rosé can age).
One could also argue that there seems to be a weak negative correlation between perceived quality and when it was tasted in the line-up (i.e., that later tasted wines generally scored lower), while that certainly is a concern, statistically speaking it did not reach the 0.3 threshold, so it is dismissed as little to no correlation.
Aren’t you sorry you did not pay more attention in that Stats class in school?
A couple of weeks ago, seven of us tried 68 American Rosés on Saturday, trying to find the best. As promised, I am publishing my actual notes from the tasting, which we tasted in 17 flights of four wines. Here are flights 13-15.
2017 Wilakenzie Estate Pinot Noir Rosé, Willamette Valley, OR: Retail $23. Light pink. Strawberry and a touch of funk. Good fruit, moderate acidity, really nice. This is perhaps the best one in a while. Outstanding. 90-92 Points.
2017 Onward Rosé of Pinot Noir, Redwood Valley, CA, Hawkeye Ranch: Retail $22. Really pale. Closed on the nose with not much fruit, but on the palate, I have to say, there is some subtle complexity here. This might need some time. Now? A silent killer, in a few months? Could be a cold-blooded assassin. Very Good to Outstanding. 89-91 Points.
2017 Rodney Strong Vineyards Rosé of Pinot Noir, Russian River Valley, CA: Retail $25. Pale pink. Fruity and really nice, sweet fruit. Wonderful on the palate. Really nice, and while the finish is a tad short, this is close to gangbusters. Outstanding. 90-92 Points.
2017 J. Bucher Rosé of Pinot Noir, Russian River Valley, CA: Retail $25. A touch of menthol on the nose, but surprising weight on the palate, with rich fruit, good tartness, and a relatively short, but vibrant finish. Very Good to Outstanding. 89-91 Points.
2017 Scheid District 7 Rosé, Monterey County, CA: Retail $15. Sweet strawberry nose with an ounce of funk. Peachy and sweet, with fairly good acid, this is a delightful wine. Very Good to Outstanding. 89-91 Points.
2017 Chehalem Three Vineyard Rosé, Willamette Valley, OR: Retail $25. Tight nose with some citrus. Round and lush with good fruit, just missing the tartness. Very Good. 87-89 Points.
2016 Passaggio Tempranillo Rosé, Clarksburg, CA: Retail $32. Really light in color and fairly closed on the nose. Eventually, a bit of red fruit. Also closed on the palate, but after a bit of time en bouche it began to sing: complex flavors and above average finish. Nice. Very Good to Outstanding. 88-90 Points.
2016 Benovia Rosé of Pinot Noir, Russian River Valley, CA: Retail $32. Sweet red fruit on the nose, which continues onto the palate. Very nice flavors, with a lingering finish. Very Nice. Very Good to Outstanding. 88-90 Points.