Monday, June 18, 2018

Not all retailers cherry-pick the best critic scores for their wines

I noted in an earlier post that wine retailers often cherry-pick the critics' quality scores that they quote when advertising their wines (Laube versus Suckling — their scores differ, but what does that mean for us?). They can do this because the quality scores from different critics often do not agree, and there is a pecuniary advantage to the retailer for any given wine to appear to be as good as it can be.


However, not all wine advertising is like this. For example, Wine Rack Media, a company offering services to online wine retailers, makes available the scores (for any given wine) from a range of critics, and these scores are sometimes quoted in full by retailers. For example, both Fisher's Discount Liquor Barn and Incredible Wine Store have web pages "powered by Wine Rack Media", and they both use that company's wine database.

Let's take a specific example of a well-known wine, to see what information we get: Caymus 'Special Selection' Cabernet Sauvignon. Checking online, we are given wine descriptions (including quality scores) for most of the vintages from 1978 to 2005, many of them from multiple sources: Wine Spectator, Wine Enthusiast, Wine and Spirits, Stephen Tanzer, Connoisseurs' Guide to California Wine, and Vintage Tastings.

What is more interesting, though, is that we are often given multiple scores from each of these sources, rather than a single score. Most reputable sources of wine-quality scores are likely to have tasted the wines on different occasions, often by different tasters, so that we do have multiple scores available. Most wine retailers report only the highest of these scores, for each wine (ie. a cherry-picked score); but not in this example.

The first graph shows all of the quality scores we are given from the Wine Spectator magazine, with the vintages listed horizontally and the scores vertically. Where there are multiple scores for a particular vintage, I have connected them with a line.

Wine Spectator quality scores for Caymus 'Special Selection' Cabernet Sauvignon

As you can see, the repeated scores tend to differ from each other by at least a couple of points, although one pair differs by 16 points (82 versus 98) and some others differ by 10 points and 9 points. Clearly, cherry-picking a score would be very effective here. Indeed, all of the scores below 90 points have a paired score that is much higher — we could, if we were so inclined, easily claim that the Caymus Cabernet is consistently better than a 90-point wine!

For comparison, the second graph shows some of the other quality scores, as well.

Several critics' quality scores for Caymus 'Special Selection' Cabernet Sauvignon

Note that the Tanzer scores are almost all lower than the other scores for the same vintage, while the Connoisseurs' Guide scores are usually lower. Indeed, the only vintages for which there is good agreement are 2000 and 2004 (where all of the scores are within 1 point).

With this sort of presentation of the quality scores we can thus see at a glance where we stand with regard to the variability among critics' scores. If only all wine retailing was this honest about the multitude of data available.

Monday, June 11, 2018

Actual wine consumption versus the recommended maximum

Many, if not most, countries have an "official" recommended maximum level of wine consumption per week for adults. However, the recommended value differs rather a lot between the countries. Moreover, the observed weekly consumption of wine also differs between countries. I have therefore wondered how these two values compare — the actual consumption versus the recommended maximum one.


For the recommended wine-intake values for each country, I have used the data from the International Alliance for Responsible Drinking (IARD), which were last updated in January 2018. The values given are in grams of alcohol per week (for each adult), with separate values for males and females, if they differ. We know that one 750 ml bottle of 12.5% wine equals 74 g alcohol, so we can use this to convert the values to bottles of wine per week.

For the actual wine-intake values for each country, I have used the Annual Per Capita Wine Consumption 2014-16 list from the AAWE facebook page. The values given are liters of wine per person per year. To convert these to "per adult", I have used the World Bank data for the percent of each population aged 15-64 years (this is apparently a standard age group for "adults"). I then converted these new values to bottles of wine per week.

This means that I now have two figures for bottles of wine per week per adult, one estimating actual consumption and one describing the recommended maximum, for a range of countries. Clearly, the value for actual consumption does not take into account what proportion of the population actually drinks wine.

The only complication is that the observed consumption values combine both males and females, whereas the recommendations for the sexes sometimes differ — where they do differ, the value for females is typically one-half or two-thirds of the value for males. Interestingly, there has been a recent trend for countries to lower the recommended limit for wine consumption by men to the same value used for women. Apparently, we are moving away from the James Bond / Humphrey Bogart hard-drinking lifestyle, at least as a medical recommendation for men (see my blog post James Bond, alcoholic).

Anyway, the data in the table show the actual consumption (bottles of wine per week per adult) as a percentage of the recommended limit for men. Most countries have a recommended limit for men of 1.9-2.3 bottles per week.

Croatia
France
Netherlands
Switzerland
Portugal
UK
Belgium-Lux.
Italy
Denmark
Austria
Australia
Sweden
New Zealand
Uruguay
Germany
Georgia
Chile
Greece
Bulgaria
Ireland
Argentina
Romania
Finland
Canada
USA
Spain
90%
90%
88%
83%
75%
74%
63%
60%
53%
49%
47%
45%
45%
44%
39%
39%
36%
36%
31%
31%
30%
27%
26%
20%
16%
15%

Note that most of the countries have wine intake that is less than 50% of the recommended maximum wine consumption. In theory, this should make the medical people happy. However, there are two ways to be at the top of this list: (i) have a high consumption, and (ii) have a low recommended maximum consumption.

The countries with lower limits than 1.9 bottles per week include Bulgaria and the United Kingdom (1.5 bottles), Chile (1.3), and the Netherlands (1.0). This explains why the UK and the Netherlands are near the top of the list, even though their consumption is not particularly high.

The other top countries are there because of their high wine consumption. Indeed, Croatia, Portugal, France and Switzerland each consume 25% more wine per adult than does their nearest rival (Italy).

The countries with the highest recommended limits include Argentina, Canada and the USA (2.7 bottles per week), Greece (2.8), Romania (3.0), and Spain (3.8). This explains why these countries occupy the bottom places on the list — they set high limits, and so their people's consumption gets nowhere near that limit.

Note that Chile and Bulgaria have low recommended limits but even lower wine consumption.

Finally, it is worth noting that those countries with wine-consumption values exceeding 50% are likely to have average consumptions that exceed the recommended value for females, since these are often half that of the values for males. This is true for Croatia, Switzerland and Portugal. Also, since the value for actual consumption does not take into account what proportion of the population actually drinks wine, there will be many people of both sexes who exceed the recommended limit, possibly by a great margin.

Monday, June 4, 2018

What is a "most admired" wine brand?

The short answer is: It depends on what year it is!

Each year, the April edition of Drinks International magazine contains a supplement with a survey called The World’s Most Admired Wine Brands. A group of people are asked to vote for the wine brands they "most admire" based on the criteria that each brand should:
  • be of consistent and / or improving quality
  • reflect its region or country
  • be well marketed and packaged
  • respond to the needs and tastes of the target audience
  • have broad appeal among wine consumers.
So, while "admiration" is admitted to be an intangible thing, there are clearly a few pointers here.


The people polled are drawn from "a broad spectrum of the global wine community", which apparently includes: masters of wine, sommeliers, commercial wine buyers, wine importers and retailers, wine journalists, wine consultants and analysts, wine educators, and other wine professionals. There were only 60 people involved back in 2012, but there are now more than 200.

The people could originally vote for up to six wine brands, but apparently they are now asked for only three choices. Furthermore, they are provided with a list of previous winners, including "a list of more than 80 well-known brands and producers, but as usual we also encourage the option of free choices". David Williams has presented his own take on what it means to take part in this poll.

The polls

I have compiled the poll results for the years 2011-2018 inclusive. Each of the published lists contains only the results for the top 50 ranked wine brands in that year — all we know about the other brands is that were ranked lower than 50th place in that year.

Across the 8 years, 98 brands have appeared at least once in the lists. However, only 15 of these brands appeared in all 8 lists, with a further 10 brands appearing in 7 of the 8 lists. There were 20 brands that appeared only once each. There is thus a great deal of variability in "admiration" from year to year.

The first graph shows the yearly data for 9 of the wine brands that appeared in all 8 lists, with the vertical axis showing their ranking from 1st on down to 50th. As you can see, even for these brands their results varied dramatically from year to year. Indeed, only the top three brands have shown any consistency at all — these are Torres (from Spain), Concha y Toro (from Chile), and Penfolds (from Australia). Indeed, Torres was ranked either 1st or 2nd every year, which must make it the most admired brand of all.

The "most admired" wine brands from 2011-2018

To this list of consistent brands we can also add E. Guigal (from France) and Ridge (from the USA), both of which missed the 2011 list but had relatively consistent ranks thereafter (mostly in the top 10).

The second graph focuses on the wine brands from Bordeaux, including those that did not make it onto all 8 lists — this is far and away the best-represented wine region in the lists, with 10% of the brands.

The "most admired" Bordeaux wine brands from 2011-2018

These are all top wine chateaus, of course, with the most expensive wines. The most successful of them seems to be Château Margaux, but even it varies in rank from 7 to 29 across the years. Château Latour and Château d'Yquem are the only ones to get into the top 5 in at least one year, but these two chateaus then missed the list entirely in other years. Clearly, Bordeaux does not engender unmitigated admiration in the wine world.

As far as countries are concerned, France hosts 20% of the admired wine brands:
Bordeaux
Rhône
Burgundy
Languedoc
Beaujolais
Provence
General
10
3
2
2
1
1
1
You will note the very poor showing from Burgundy — there are not many large wine brands (only Louis Latour makes most of the lists), but instead a host of smaller brands marketing very expensive wine (Domaine de la Romanée-Conti makes it onto two lists).

The remaining countries include:
Australia
Spain
USA
New Zealand
Chile
Italy
South Africa
Portugal
   [Porto
Germany
Argentina
Canada
China
Hungary
Lebanon
13
 13
11
8
7
6
6
5
4]
3
2
1
1
1
1
Even though the Porto region hosts 4 of Portugal's admired wine brands (Dow's, Graham's, Sandeman, Taylor's), these have only ever made it onto two of the lists (2016 and 2017).

The brands

I have provided a summary of the data relating to the individual brands in the following network. It displays all of those brands that appeared in at least 50% of the lists (ie. 4 out of 8). This is a form of multivariate data summary, as described in my post Summarizing multi-dimensional wine data as graphs, Part 2: networks. [Technical details: this is a neighbor-net based on the gower distance.]

Each brand is represented by a dot in the network. Brands that are closely connected in the network are similar to each other based on their ranks across the 8 polls, and those that are further apart are progressively more different from each other. So, for example, the three top-ranked brands (Torres, Concha y Toro, Penfolds) are together at the top of the diagram, followed by the next pair (E. Guigal and Ridge). From there, the network progresses down to the less-admired wine brands at the bottom.


Note that Guigal and Ridge are at the head of a bunch of 13 wine brands, all of which are relatively highly admired. Then there is a group of 7 intermediate brands (the network area from Oyster Bay down to Pétrus), plus Marqués de Riscal on its own — the latter is isolated in the network only because it inexplicably missed the 2017 list.

You will also note the proximity in the network of Yellowtail to Château Mouton-Rothschild and Cheval Blanc! This should make you wonder about the criteria for "brand admiration".

The basic issue with these lists

There are potentially at least three things wrong with the "best of" type of list: (i) there is rarely any clear idea of what "best" is supposed to mean; (ii) the list is of arbitrary length (eg. Top-10 or Top-50 only); and (iii) the ranking does not reflect the differences in the original scores. In this instance, we have some idea of what "most admired" is supposed to mean; but the other two issues definitely apply here.

More importantly, there is an issue with interpretation that is rarely mentioned. To say, as many of the media have, that "In the 2018 poll the industry voted Torres the most-admired wine brand" is wrong, because there is no evidence that the people polled did any such thing. Indeed, we do not know how many people actually did put Torres (or any other brand) on their own personal list of three brands. All we know is that Torres is listed as the no. 1 brand because more people put it on their list than did so for any other brand — it may have been a lot of people or it may not.


To provide a concrete example of what I mean, I will refer to my previous discussion of a similar situation for the "Greatest Films Poll" produced every decade by Sight & Sound magazine, which lists the top films as voted by selected film critics. The critics are asked to each list 10 films; and in the most recent poll Alfred Hitchcock's film Vertigo topped the overall poll. However, the vast majority of the critics (77%) said that this film doesn't even belong in the top 10 (ie. it was not on their personal list), let alone first. However, it is listed as the no. 1 film, because more critics (23%) put it on their list than did so for any other film. Similarly, 91% of the critics said The Searchers should not be in the top 10, and yet it is ranked no. 7. So, the rank order of the films is simply that — a rank order; it does not tell you how many critics think highly of each film.

This is the same point that I am making for the wine brands, although in this case I do not have the detailed information to say exactly how many people listed each brand.

In a similar vein, anyone who knows anything about banking will known that "the most trusted bank" is nothing more than the "the least mistrusted bank". The distinction is not trivial for the consumer.

This knowledge does not stop the misuse of these sorts of lists, of course. For example, Voxy recently noted: "Villa Maria [was] named the world’s most admired New Zealand wine brand by Drinks International". This is, strictly speaking, true, but Villa Maria did go down from 4th rank overall last year to 8th this year!

In a future post I might looks at some of the other industry awards, such as The World Ranking of Wines and Spirits and The Drinks Business Awards.

Monday, May 28, 2018

Growing degree-days maps of Australia and the western USA

In a previous post (Wine is sunlight held together by water) I discussed the use of both the Winkler Index and the Köppen-Geiger Climate Classification for determining suitable locations for high-quality wine-making vineyards. At the global scale the Climate Classification is probably the most suitable tool, but at more local scales the Winkler Index comes into its own, which is what I will discuss here.

Albert J. Winkler

This Index is based on Maynard Amerine and Albert Winkler's idea of Growing Degree-Days (GDD) — the sum of the daily temperatures above 10 °C (50 °F) during the grapes' growing season. There are several ways that the data necessary for the calculations could be collected for any one spot on Earth, but a standardized version was developed some time ago by Gregory V. Jones and his colleagues. They have published several papers about various regions of the world:
The basic idea is to take digital maps (latitude, longitude, and altitude) and digital weather records (monthly maximum and minimum temperatures, plus precipitation), and combine the two using a computerized climate model. The output is a grid map, with an estimation of the GDD value for each cell in the grid. The cells can be of any chosen size, but a common choice is 1 km square (100 hectares or 250 acres). Obviously, this is not a fine enough scale to tell anyone where to plant their vines, but it is good enough to tell them whether they are looking in the right area or not.

Of interest to us here is the fact that the above publications each contains a gridded map showing the Winkler Index zones for a different part of the globe. I have reproduced here a version of the maps for the western USA (California, Oregon, Washington, and Idaho) and for Australia.

Growing degree-days map of vineyard land in the western USA

The map of the western USA is based on 400x400 m cells = 16 ha = 40 acres. The colors refer to the five Winkler Index zones (I-V with increasing heat summation). Most of the northern areas of the map are unsuitable for viticulture because the heat summation is too low (<850 degree-days), while the south-eastern parts have a heat summation that is too high (>2700 degree-days). The Central Valley of California is clearly indicated in red, being the hottest place with grapevines in the western USA.

This map can be compared to the boundaries of the American Viticultural Areas (AVA), which would indicate which grape varieties might be most suitable for each AVA. However, the spatial variability of the GDD values within any one AVA can be large. For example, the best-known AVA, Napa Valley, mostly lies within Region III, but it actually ranges from Region I to the lower part of Region V. The most important cause of this effect throughout the western USA is variation in altitude within the AVAs. Consequently, parts of most AVAs have never been planted with grapes, and probably never will be.

Growing degree-days map of vineyard land in Australia

The grid cells of the Australian map are 25 sq.km each, since the map covers the whole continent. Note, however, that very little of Australia is actually suitable for viticulture, unlike North America or Europe, because the heat summation is way too high. Only in the south-eastern part are vines absent because the heat summation is too low.

Most of the suitable vine-land area of Australia already has formal Geographic Indication (GI) viticultural areas, which are outlined in red on the map. Those areas with suitable GDD values but that do not (yet) have vines are, in fact, almost all too dry for viticulture.

For Australia, the biggest influences on the differences in GDD values between GIs are latitude and continentality. For example, note that the GDD values generally decrease from north to south down the east coast, with the coolest climates being on the island of Tasmania. Furthermore, the inland GIs are more likely to have greater index values than are GIs closer to the ocean. However, the latter effect is interrupted by the Great Dividing Range, which runs all the way down the east side of Australia. The upland areas are much cooler than are the coastal areas. Indeed, the northern GI areas are all at high altitude.

As noted above for the western USA, even within each of the GI areas, parts have never had grapes and probably never will have, due to variation in altitude. This is particularly important in those GIs located in the Great Dividing Range.

Both Australia and the western USA have areas within all of the Winkler Index zones. For example, Jones et al. list some comparable Australian GIs and western US AVAs, based on their median GDD values:
  • Region I – Southern Tasmania (GI) with Puget Sound (AVA) and numerous Willamette Valley sub-AVAs
  • Region II – Coonawarra (GI) with Red Mountain (AVA); Upper Goulburn (GI) with the Sonoma Coast (AVA)
  • Region III – Bendigo (GI) with Alexander Valley (AVA) and the Dry Creek Valley (AVA); Margaret River (GI) with the Napa Valley (AVA)
  • Region IV – Adelaide Plains (GI) with Lodi (AVA)
  • Region V – Swan District (GI) with Madera (AVA).
Finally, it is worth noting (as Jones et al. do) that viticultural areas with similar grape varieties and wine-making reputations do not always have the same GDD values. For example, Burgundy (France), the Willamette Valley (USA), and the Yarra Valley (Australia) each excel with Pinot noir and Chardonnay, but the Yarra Valley (median 1558 degree-days) is significantly warmer than is either Burgundy (1118) or the Willamette Valley (1081). Similarly, Bordeaux (France), the Napa Valley (USA), and Coonawarra (Australia) are often compared based on the quality of their Bordeaux-mix of red grapes, but the Napa Valley (1883) is substantially warmer than is either Bordeaux (1410) or Coonawarra (1457). Intriguingly, Margaret River, a GI with a rapidly rising reputation for producing Australia's best Bordeaux-mix wines, is remarkably similar to the Napa Valley (ie. median 1844 degree-days).

Conclusion

There are a lot of limitations of using Growing Degree-Days to assess areas for their climatic suitability for growing high-quality wine grapes, but nevertheless they can be very informative, especially for indicating unsuitable locations.



PS. This week marks the second anniversary of this blog.

Monday, May 21, 2018

Forecasts, predictions and guesses in the wine industry

I read an online article recently that used the words "guesses", "forecasts" and "predictions" in the same paragraph, as though they were interchangeable words. However, if these words all meant the same thing then at least two of them would be redundant. So, we may safely conclude that there must be some difference(s) between them, however subtle.

The point of this post is that the differences between them are not subtle, at all. Indeed, their differences are important to our success when making decisions about the future, although writers seem to be rather confused about the difference.* They criticize the forecasters because their "predictions" are not accurate — however, the pundits did not make predictions, they made forecasts.


A helpful analogy

For the purposes of this blog post, imagine one of those old-fashioned collections of pigeonholes, in which hotel keys and letters are stored (see above). Furthermore, imagine that each one of the pigeonholes represents a piece of time, and that the events of that time are written inside a single envelope within each pigeonhole. Due to "time's arrow", as the physicists call it, we cannot open these envelopes in any arbitrary order — we must proceed sequentially through the pigeonholes (perhaps starting at the top-left of the collection).

At any given moment, we can read the contents of one pigeonhole only — we call this the "present". We have opened all of the envelopes leading up this one envelope — we call this the "past". We cannot go back and read those envelopes again, but instead we have to remember their contents, perhaps relying on our own memory, or perhaps we wrote a few brief notes at the time. All of the remaining envelopes remain unopened — we call this the "future".**

In spite of our lack of knowledge, we are interested in the contents of the unopened envelopes, because that would tell us what will happen to us next. Perhaps we don't want too many details about some of those envelopes, because one of them tells us when, where and how we will die — that envelope may be best left sealed until we get there!

However, the other envelopes contain information that it would be valuable to know beforehand. For example, in the wine industry they will tell us about future grape-growing seasons, about future wine-making techniques, about future wine-consumption trends, and about future marketing and selling opportunities. A peek into some of those envelopes would make a lot of people rest more easily at night.

But, since we cannot peek, we instead use guesses, predictions or forecasts — we hold the envelopes up to the light, and we try to read what is written inside.


Guesses

Anyone can make guesses. Every time we call heads or tails when flipping a coin we are making a guess. So, we all understand guesses, because we can all do it.

Sometimes we make what we call an "educated guess", which simply means that we have some previous information that might improve the success of our guessing. However, even this does not involve any formal procedure for making the guess — we merely modify our guess in some subjective way, based on what we already know.

There is nothing intrinsically wrong with guessing — after all, we will be right some percentage of the time. Indeed, the old finance "40% Rule" says that one can look good simply by following any proposition with a 40% probability (see How do pundits never get it wrong? Call a 40% chance).

The basic issue, however, is that guesses almost always involve either optimism or pessimism — people guess the outcome that appeals to them at the moment, either because they want some particular thing to happen or they want to stop something. There is too much unhelpful emotion in guessing.


Predictions

Predictions might lie at the other extreme from guesses, in that there is supposedly an explicit method being used to produce the prediction. However, the method of producing the prediction is often not revealed, so that exactly how predictions differ from guesses is not always clear.

A prediction simply declares: "This is what will happen next." This approach leaves the door open for charlatans to make predictions, just as much as it does for people with a serious method of prediction. The world of illusions masquerades as "magic" solely based on this principle. Believing in magic is like believing in Santa Claus — it is fun as a form of entertainment, but it should not be taken too seriously (except by the very young).

This, then, is the basic issue with predictions, that there is no known method of making them repeatable (intuition, crystal balls, tarot cards, consulting prophets, etc). I can predict the weather in any way I like, just as I can predict the outcome of tomorrow's football match, whether or not I tell you how I am doing it.


Forecasts

The difference between prediction and forecasting is pretty simple. Forecasting says: "If things continue the way they have in the past, then this is what will happen next." Prediction leaves out the caveat, and simply declares: "This is what will happen next." For example, we cannot forecast the result of a single toss of a coin, but we can forecast the overall outcome of a long series of tosses; on the other hand, we could predict the result of a single toss of a coin.

So, forecasts involve the relative probabilities of future events, whereas predictions are usually presented as being more certain. Forecasters do not usually eliminate alternative possible futures, they simply claim that they have small probabilities of occurrence. Dealing with a probabilistic world is not always easy for people (see The best way to think about probabilities).

So, technically, "weather forecasting" is not the same as "weather prediction"; and the various weather bureaus around the world insist that what they are doing is forecasting the most likely weather not prediction the future. They do not have a crystal ball, just a bunch of equations. In a previous post, I noted that people are actually smarter than the skeptics think. We keep records (ie. we describe the world), we think about the patterns observed in those records (ie. we explain the world), and we work out how we might respond (ie. we try to forecast the future). So, weather forecasting is not necessarily a futile exercise, as many people have suggested (see my blog post on Foretelling the weather.)

Forecasts often involve some mathematical model — we feed information about the present and the past into the mathematical equation(s), and a forecast comes out the other end. As an example, in my post on Getting the question right, I presented a couple of simple economic models, and used one of them "to explore possible forecasts for future [wine] export growth."

The basic downside to forecasts is that the future is often disconnected from the present and the past. The forecasters have only the present (which we can measure) and the past (which we have recorded) — if either of them turns out to be irrelevant, then the forecasters don't have any more ability than anyone else to talk about the future. There are "unforeseen circumstances" that can arise, and when they do all, forecasts go out the window. For example, tomorrow's weather forecast is unlikely to be correct if a volcano erupts overnight. If you treat a forecast as a prediction, then you are likely to be disappointed in the outcome (for an example, see How the oil rally took forecasters by surprise).

Furthermore, different mathematical models can produce different forecasts, so that there is much gnashing of teeth during any discussions about the appropriate model to use. This is at the heart of many discussions about Global Warming, for instance. To return to my analogy, changing weather patterns are self-evident when we compare today's envelope to the ones we have opened in the past, but it is not always obvious how to use this information to think about the unopened ones.

I noted in a previous post that forecasts are problematic no matter how they are implemented, so that the modern reliance on Artificial Intelligence (AI) will not necessarily help. Nevertheless, the AI forecasts are expected to improve through time, as more and more data are gathered, and the AI system continually adjusts itself based on newly found patterns in the data (ie. they take the newly opened envelopes seriously).


A few wine-related examples

(i)
Consider the following forecasting example, which is of interest to grape growers in California (see Climate scientists see alarming new threat to California).

The most recent drought in California was caused by a high-pressure system in the north Pacific. A low-pressure system circulates air clockwise, while a high-pressure system circulates air anti-clockwise. Therefore, north-Pacific low-pressure systems direct the flow of moist tropical air towards California, while high-pressure systems direct that air toward Alaska, instead. So, with a low-pressure system it rains in California (and there is a drought in Alaska), and with a high-pressure system it rains in Alaska (and there is a drought in California).

For the wine business, then, the important forecast is whether the future holds the prospect of more high-pressure ridges in the northern Pacific. The current forecasting models say that, if the polar ice-cap continues to melt, then: "yes".

Note the three components of the forecast: (i) the "past", which consists of observations of the behavior of high- and low-pressure weather systems; (ii) the "present", which consist of observations of melting ice up north; and (iii) the "if", which relates to whether the past and current situations will continue into the future. Points (i) and (ii) refer to envelopes that we have already opened. Point (iii) is the basic assumption of forecasting that makes it different from a guess or a prediction.

(ii)
Another recent example of forecasting is in the new book edited by Kym Anderson and Vicente Pinilla (2018) Wine Globalization: a New Comparative History (Cambridge University Press). This book (and its companion data volume) provides an extensive coverage of all sorts of wine-related data from 1835 to 2015, intended to illustrate time trends in the wine industry as a whole. This is ideal base material for forecasting, because it quantitatively lists the contents of all previous envelopes, plus the present envelopes. Here, we can genuinely say that if things continue the way they are, then the immediate future can reasonably be forecast; and this is what happens in Chapter 18 of the book (by Kym Anderson and Glyn Wittwer: "Projecting global wine markets to 2025").

Mind you, the book also illustrates the potential folly of trying to extend forecasts too far into the future — after all, even weather bureaus only produce 10-day forecasts. Things do not "continue the same" for long, and the wine industry in every country provides good examples of ups and downs. For instance, the European Union once dominated world wine exports (even as recently as 1988 it accounted for c. 80% of exports), but it no longer does so (<50% in 2014) — if we want to extend this downward trend indefinitely, there must come a time when the EU has no wine exports at all!


(iii)
Finally, a much earlier example of forecasting was the attempt in 1990 by Orley Ashenfelter to forecast the future quality of Bordeaux red-wine vintages, based on observed relationships between vintage quality and summer temperature. In a previous blog post (Fifty years of Bordeaux vintages), I noted that his forecast for the 1989 vintage was spot on, while his forecast for the 1986 vintage was way off. This inconsistency answers the question recently posed by Peter R. Orszag: "An economist’s method predicts a vintage’s quality using only statistics. Why hasn’t it caught on?" More consistency is needed, that's why.

Issues with forecasting

I cannot finish this post without mentioning one further characteristic of forecasts that we all need to be aware of — some events are consistently much easier to forecast that are others. I once wrote a whole blog post about this, using sports forecasting as my example: Forecasting and predicting sports results. [Please ignore the fact that the post claims to have been written by someone named "Christina Pikas" — the file has been corrupted, but that blog is now defunct and I cannot get it fixed.]

The point of the example is that it is easier to forecast sports events that are played until a specified goal is reached, compared to sports events that terminate at a specified time. For example, a tennis match proceeds until someone wins two sets (or three in some competitions), while a basketball game ends at the moment the buzzer is sounded. In the latter case, the game situation even one second before or one second after the buzzer is immaterial, and therefore forecasting the situation precisely at the buzzer is very hard. In one case (tennis) you are forecasting general superiority of the players, and in the other (basketball) you are forecasting a specified moment in history.

Keep this distinction in mind when you are evaluating a forecast — what chance did the forecaster actually have of ever getting it right?

Conclusion

Don't take a guess, and don't make predictions. Forecasts do have their limitations, but they can also do a lot for us.



* The title of the best-selling book by Ian Ayres, Super Crunchers: How Anything Can Be Predicted, is a classic example of confusing the words "prediction" and "forecasting". The book is an interesting discussion of using Big Data for forecasting, but it never once mentions oracles (ie. predictions).

** This idea comes from Fred Hoyle's novel October the First is too Late.

Monday, May 14, 2018

11 tasters and 20 wines, and very little consensus

In my previous post about wine-quality scores (Why comparing wine-quality scores might make no sense), I outlined why it is often (usually?) a bad idea to mathematically compare the numbers — even when the scales look the same, the scoring systems are probably not the same, so that we cannot tell what any similarities or differences in scores actually mean.

I also provided a list of blog posts in which I have previously illustrated just how diverse are the different scoring schemes among critics. What I have not done, until now, is discuss the best example that I know of, where it is clear that many members of a single group of wine tasters had a different idea of what the wine-quality numbers should mean.

The tasting I am referring to is what has become known as the Judgment of Paris. I have previously provided an introduction, for anyone not familiar this event (A mathematical analysis of the Judgment of Paris).


My interest in this tasting actually has nothing to do with its social interest or importance — here, I am solely interested in the fact that the data highlight almost all of the mathematical problems with evaluating wine quality using numbers. Nominally, the 11 wine-tasters were using the same 20-point wine-quality scale, but in reality there was a different scoring scheme for each person — that is, there are clear differences in their interpretation of the 20-point scale.

For us, this means that there was a lack of repeatability among the wine-quality assessments, and therefore severe problems with taking average quality scores.

Background

The tasting was organized by Steven Spurrier and Patricia Gallagher in 1976. Its fame derives from the fact that it (deliberately) coincided with the US Bicentennary. Its infamy derives from the fact that several people with a modicum of mathematical knowledge have been criticizing it ever since then (these are listed in the blog post referred to above).

The situation is this: 9 French people from diverse wine backgrounds were invited to a tasting of French and Californian wines, 10 cabernet-based and 10 chardonnay-based. All 11 people (including the organizers) were asked to score the wines on a 20-point scale, without further instruction — they were all blind to the origins of the wines (including Spurrier and Gallagher).

The outcome was a set of scores based on whatever scoring system the taster chose (ie. their own personal criteria), but their results were all presented on the same scale (ie. a number from 0-20). There is thus no reason whatsoever to assume that the numbers are comparable, unless we can demonstrate that the tasters were all using the same scoring system.

There is one technical difficulty concerning the data for the white wines, which I have previously presented (Why we no longer have the data from the Judgment of Paris). For our purposes here, I have left the scores as originally reported, even though the numbers do not add up to the reported totals.

The data

I am not interested here in evaluating either the wines or the people, so that both will remain anonymous for most of this blog post. I have presented the original data in the first pair of graphs, with the scores shown vertically for each taster (horizontally). Each wine is represented by either a + or an x, so that if two wines were given the same score by the same taster then the result looks like *. Note, also, that the tasters are not in the same order in the two graphs, but are simply listed in the order of decreasing average score for that wine type.

Scores for the Judgment of Paris cabernet wines
Scores for the Judgment of Paris chardonnay

Two things are immediately obvious: (i) some assessors scored higher than others, and (ii) some assessors used a much greater range of scores. The mathematical consequences of these characteristics will be explained below.

For the moment, though, let's look at some of the features of the data. For example, the top score for the red wines (17) was awarded by five different people, although it was given to four different wines. So, there is a ceiling for the upper score but not for the lowest score, which varied greatly between tasters (from 2 to 8).

Labeling the tasters A-K (in alphabetical order), then we can say the following: (a) at one extreme, G and K produced one of the highest or lowest two scores only 3 times, while (b) at the other extreme, E and J produced one of the highest or lowest two scores 13 times out of 20 wines. That is, some of the tasters used low scores quite consistently.

If we were to drop the top and bottom scores for each of the 20 wines (ie. 40 scores), as is often done for sporting competitions where value judgments are involved, then: (a) at one extreme, B would never have a point dropped, and G would have only one dropped (for being the highest), while (b) at the other extreme, J would lose at least 6 out of 20, and C and D would lose at least 5 out of 20 scores.

These patterns can be summarized as shown in the next graph, where each point represents one of the assessors, with their average score shown horizontally and their score range (maximum minus minimum) shown vertically.

Average scores versus the variability of the scores for the Judgment of Paris

Clearly, some tasters scored consistently high (those at the bottom-right of the graph), while others used a broad range of scores (at the upper-left), which must consequently lower their average score. The two non-French tasters are shown in pink. Obviously, including or excluding these two assessors can be expected to make a difference to the calculation of the average score for any given wine.

Finally, we can try to summarize the whole thing using a multivariate ordination. This technique is described in a previous blog post (Summarizing multi-dimensional wine data as graphs, Part 1: ordinations) — in this case, the method tries to put the tasters in some order based on their scoring patterns. This order is shown in the final graph, where the tasters are named — at the top are those people who scored highest, with the lowest (ie. using a lot of the score range) at the bottom.

Multivariate ordering of the score patterns for the tasters at the Judgment of Paris

There are five people whose scoring schemes were rather similar to each other, clustered in the middle of the order — Claude Dubois-Millot and Aubert de Villaine are at what is called the centroid (the weighted average), making them the "middle-scoring" tasters. The other six people deviated more or less strongly in their scoring schemes, from this middle ground of five people, with some of them diametrically opposed to each other in their behavior.

What does this all mean?

So, what is the importance of this obvious variation in scoring schemes among the tasters? Mathematically, this has an enormous influence on calculating average scores for the wines. In effect, those people who use a large part of the scoring scale will have a disproportionate effect on the average score.

Consider this simple example of two people scoring two wines on a 20-point scale:

Taster A
Taster B
Average
 Wine 1
 16
 18
 17
 Wine 2
 18
 14
 16
The two tasters disagree as to the relative quality of the two wines (A prefers 2 and B prefers 1). However, Taster B uses a greater range of scores than does Taster A, and consequently the average scores for the wines reflect Taster B's preference, rather than Taster A's — using a low score greatly down-weights the average score for that wine.

So, the people who had the biggest effect on the average scores of the Judgment of Paris wines are those listed at the bottom of the final graph above — and the further down the list they are then the bigger was their effect.

Conclusions

We cannot possibly work out whether the tasters agreed about the wines or not — each person had their own scoring scheme (although five of them were similar), and so we cannot work out how to interpret their scores. And yet, their scores were averaged, and then presented to the world as having some meaning about the quality of the wines. This was mathematical nonsense, because it allows some of the people to have a disproportionate effect on the outcome.

Not unexpectedly, the results of this Paris tasting were not repeatable at the time (see Was the Judgment of Paris repeatable?), least of all by those people who used an objective wine-quality scheme rather than a subjective one (eg. the Vintners Club, where all of the tasters must use the UCDavis scoring scheme — Did California wine-tasters agree with the results of the Judgment of Paris?).

Clearly, if everyone scores identically (ie. uses the same scoring scheme) then there would be no problem (as at the Vintners Club); and if everyone scores differently there would be little purpose to averaging the scores. At a tasting like the Judgment of Paris, the obvious procedure is that everyone must get together beforehand and agree about which scoring scheme will be used, as happens at many modern-day wine shows.

The only alternative is to try some fancy mathematics on the scores, to try to make them comparable, before averaging them. I will write about this idea in a later post.

Monday, May 7, 2018

Wine monopolies, and the availability of wine

To some people, the concept of a government-controlled monopoly is like a red rag to a bull — basically, they "abhor governmental pricing and entry controls as manifestly causing waste and inefficiency, while denying consumers the range of price and service options they desire" (Paul Stephen Dempsey).

For example, we could cite the example of Santiago (in Chile), which turned its transport system from a commercial one that earned a profit of $60 million per year into a government-controlled one that now loses $600 million per year, instead (see Planning order, causing chaos: Transantiago). However, we should not conclude from this that all public transportation industries are universally reviled as either wasteful or inefficient (cf. trains in Europe or Japan).


It should not come as a surprise that a number of government-regulated locales exist in the wine industry, as well, notably in the USA (eg. Pennsylvania Liquor Control Board), Canada (eg. Liquor Control Board of Ontario), Norway (Vinmonopolet) and Sweden (Systembolaget).*1 Similarly, there is no need to assume that pricing and availability of wines in any of these regulated environments is in any way inadequate, for the consumers; but this is definitely worth investigating.

In this blog post, I present a specific example of a wine monopoly at work. I first encountered wine retailing as a consumer in Australia, where liquor retail licenses are issued by the government but there is otherwise little interference; and I now live in Sweden, where liquor retail is government-owned (but not trade sales), although Systembolaget is operated as an independent retail entity. As a consumer, I have therefore seen both types of systems.

My intention here is not to decide whether a monopoly is "good" or "bad", but simply to describe a monopoly at work. It seems to me that I am no worse off in Sweden than I was in Australia, but the systems definitely work differently.*2

Where Swedes get alcohol

The Swedish Association of Alcohol and Drug Information (Centralförbundet för alkohol- och narkotika-upplysning) estimated that, during 2016, Swedes acquired alcoholic beverages in the following ways:
Systembolaget
Traveller imports
Restaurants
Low-alcohol beer
Smuggled drinks
Home-made
Internet shops
63.2%
12.5%
10.8%
5.5%
4.9%
1.8%
1.3%
Low-alcohol beer is available for sale in supermarkets, etc, while all stronger drinks (>3.5% alcohol) are available for retail sale only in Systembolaget's shops. Trade sales (restaurants, caterers, bars, etc) can be made direct from the importers (remember, Sweden makes very little wine of its own). There is free trade among all countries of the European Union, so that internet sales within the Union are straightforward. Traveller imports have dropped from 25.7% in 2004, given that there is no longer duty-free travel within the European Union (with certain specified exemptions).

So, we can see that Systembolaget is responsible for nearly two-thirds of alcohol acquisition within a country of c. 10 million people, more than 75% of whom are of drink-purchasing age (>20 years; drinking age is >18 years). That's a lot of customers for a single retail entity, with a reported alcohol consumption of 9 liters per adult per year. For 2017, Systembolaget reported net sales equal to US$ 3.4 billion, with a 1% profit (which goes to the government).


Sytembolaget

I have written before about the fact that Systembolaget manages to sell mid- and high-quality wines more cheaply than in Britain (Why is wine often cheaper in Sweden than elsewhere?), precisely because it is government-regulated, and therefore the profit margin is controlled by legislation.*3 Indeed, the stated objective of Systembolaget is: "To minimize alcohol-related problems by selling alcohol in a responsible way, without profit motive."

Here, I discuss the range of beverages available to Swedes, which is at least as wide as most other places — this results from the sheer size of Systembolaget, along with Sweden's membership of the European Union. In 2017, Systembolaget reported selling (in millions of liters):
Spirits
Wines
Beer (>3.5%)
Cider and mixed drinks
Alcohol-free drinks
19.2
199.5
242.0
19.0
3.4

Systembolaget has c. 438 shops throughout the country, plus another c. 500 agencies where products can be ordered and received.*4 So, no Swede lives very far from a service point. This is important because, while home delivery is available in a number of places for a fee, there is no delivery cost if you collect the product yourself — you simply ask for it to be delivered to your nearest pickup point. There is internet ordering available for all products, except for the special situation where a product is still available in a particular store but is no longer available in the central warehouse — you need to ask the staff in your store to have that product transferred from the other store (for free).

When foreigners enter one of these stores, they sometimes make comments about the supposed lack of diversity among the wines available. For example, see Dan Berger's melodramatic Buying wine in Sweden — 'a vinous wasteland'. However, this is like going into Costco to find out what wines are available in the USA! There are, in fact, c. 17,500 alcohol products available in Systembolaget in any given year.

In the case of Systembolaget, the majority of the interesting wines are not to be found on the shelves (called the Ordinarie sortiment), at least not for long, and not in most stores. Wines (and beers & ciders) are released when the new vintages become available, and they can disappear quite quickly. Below, I explain the current retail arrangements, in the interests of illustrating how a monopoly can work.


Current arrangements

First, I will just list the different categories of alcohol availability for Swedes, and this will be followed by some discussion of the important differences between them.

Ordinary assortment
Occasional assortments:
    Small quantities
    Web releases
    Seasonal releases
    Local and small-scale
Order assortment
Private import
Auctions
EU ordering
Ordinarie sortiment
Tillfälliga sortiment:
    Små partier
    Webblanseringar
    Säsongslanseringar
    Lokalt och småskaligt
Beställningssortiment
Privat import
Dryckesauktioner
 

Ordinary assortment

The Ordinarie sortiment refers to the products that are stocked in the Systembolaget shops. There is a subset that is in every shop (called the Fast sortiment), plus a collection from which each local manager can choose the rest of their store's stock (depending on the local clientele). All of the products have been tasted and approved by Systembolaget's staff, who provide taste descriptions, along with suggestions for accompanying food.

At the time of writing, there are 2,565 products in this assortment. If a particular product is not in your local store, then it can be ordered from the central warehouse — delivery can take up to 4 working days, depending on where you live (for me, it is 2-3 days). The assortment is updated 4 times per year, with additions, deletions, and new vintages (or other releases). The suppliers are expected to have sufficient stock of each product to last at least 9 months.

What's in the shops is often the mass-market products, for those people who want the standard stuff, the cheapest stuff, or who are making impulse purchases. For example, the two biggest-selling whiskies do not come from either Scotland or Kentucky, but are actually both Canadian grain whiskies — and you will find box-loads of both of them stacked in every store in Sweden. If you want the other commonly known whiskies, then they are there, but you have to look a bit harder.

So, yes, if you insist on drinking your wine the same day you buy it, then what's in the store will limit you; but if you are prepared to place an order ahead of time, then you can do much, much better in terms of choice.


Small quantities

The Små partier refers to collections that are released 20 times per year (c. 60-90 products per release). The products will appear only in a selected few shops, depending on the quantities available, but they can all be ordered from the central warehouse (which is what I usually do, since my local store is not one of the selected few). How long the products remain available after release is determined entirely by the quantities —once they run out, that is it (so I always order within an hour of the release, just to be safe).

This is the assortment from which I get most of my wines, because it is here that the interesting stuff appears. What is available varies from year to year, depending on what the importers can lay their hands on. The 1,605 products released during 2017 consisted of:
Red wines
White wines
Other wines
Spirits
Beer and cider
684
341
149
112
319
The "other" category consists of rosé, sparkling, sweet and fortified wines.

If we use the price categories defined by Thach et al. (see How many wine prices are there?), then the numbers of wines were as follows:
Everyday wines
Better wines
Premium wines
Affordable Luxury
Luxury wines
Icon wines
Dream wines
0
402
1,038
224
27
0
0
The Everyday wines (< US$ 10) appear in the Ordinarie sortiment, with the Premium wines (US$ 20-50) being concentrated in the Små partier. Most of the best QPR (quality-price ratio) wines are among the Better wines (US$ 10-20).

Web releases

The Webblanseringar assortment refers to small (c. 2-30 products), and usually expensive, releases, which occur up to 35 times per year. The quantities are usually very limited (c. 6-800 bottles), and there may therefore be restricted allocations — indeed, you may miss out if you are not quick off the mark on the release date.

I can never afford these products; but during 2017, this is what was released:
Wines
Spirits
Beers
306
133
4
Using the same price categories as above, the wines consisted of:
Everyday wines
Better wines
Premium wines
Affordable Luxury
Luxury wines
Icon wines
Dream wines
0
1
17
86
186
16
3
As you can see, these are usually the more expensive releases — Luxury wines (US$ 100-500), Icon wines (US$ 500-1,000) and Dream wines (> US$ 1,000).

Seasonal releases

The Säsongslanseringar consist of 4 annual releases: Easter beer, Octoberfest beer, Christmas drinks, and Beaujolais nouveau. Swedes love their beers, and there are special brews released for different celebrations. The Christmas drinks include (in addition to beer): glögg and svagdricka. All of the seasonal drinks appear in most of the shops, or can be ordered from the central warehouse. The products available can vary from year to year.

Local and small-scale

The Lokalt och småskaligt releases occur 13 times per year. The products are available only in the stores within the local area of the producer (eg. Swedish wineries, breweries and distilleries) or of the importer. In other areas, you may or may not be able to order the products from the central warehouse. Once again, the products available can vary from year to year.


Order assortment

The Beställningssortiment refers to products that are not available directly from Systembolaget itself. These products are stored by the importers, in their own warehouses, and are made directly available to the catering trade (the Restaurangsortiment). The rest of us can often order them, too, but we have to do so via Systembolaget (and we may have to buy a case, as the minimum quantity, rather than single bottles).

At the time of writing, there are 12,935 products in this assortment. Delivery to your local store is free, and can take up to 8 working days, depending on the activity of the importer. The assortment is updated at the collective whims of the importers and their suppliers; and the online catalog is not always accurate as to the current availability. I get quite a few of my wines from this assortment — for example, if I read an interesting overseas wine review, then the wines may be available in the Beställningssortiment.

There are a couple of hundred importers in Sweden, bringing in stuff from everywhere on the planet. There is An extensive list of importers of wine to Sweden, plus a Sweden - Wine Importers & Distributors Database (which has verified information about 198 wine importers and distributors), and a full listing of the 50 biggest suppliers to Systembolaget (based on dollar value, not volume).

Private import

Since we ordinary citizens do not have import licenses for alcohol, Systembolaget will import things for us — that is, it will order any alcohol product from anywhere, provided we can give them the necessary contact details of the supplier. Systembolaget will charge a 20% mark-up for doing this, plus all of the associated transportation costs, import duties and other taxes. So, this is not necessarily a particularly economic procedure for buying anything.

Confusingly, there are also products available from the Swedish importers that they have not listed as part of their Beställningssortiment. These products can usually be ordered, via a formal Privat import request. We will almost always have to buy a case, not single bottles; and this process can take anything up to a month; so I rarely use it.


Auctions

Since retail sales of alcohol can occur only through Systembolaget, alcohol auctions need to involve this company. As a result, Systembolaget have teamed with Bukowskis auction house, to hold online auctions 8 times per year. The only auction that I have followed had c. 700 lots, none of which could be described as bargains. I may report on this in a later post.

EU ordering

There is a free-trade agreement throughout the European Union, and so alcohol is available to Swedes from anywhere within the EU.*5 Mind you, this right was only formally established by a landmark case in the EU Court of Justice, as from July 2007 — as an outcome of the prior case of Rosengren et al. in the Swedish Supreme Court (Högsta domstolen). According to my records, I have ordered boxes of wine 6 times from the UK, 11 from Italy, 4 from Germany, 2 from Denmark, 2 from the Netherlands, and once from Spain. The main limitation for me is the transport costs — many courier companies treat Sweden as being the end of the earth, and charge accordingly.

In addition, as I have noted in a previous post, wine can also be sold privately on many of the mainland European eBay sites (eg. in Austria, Belgium, France, Germany, Italy, Netherlands, Spain). The wine can then easily be sent anywhere within the EU. Indeed, many European wine shops use eBay as one of their online portals. According to my records, I have have now bought wine from eBay sellers 96 times (mostly from Germany, because of the transport costs for single bottles).

Incidentally, the EU also has allowances for bringing in duty-free alcohol from outside the EU — 1 liter of spirits or 2 liters of fortified wine or sparkling wine, plus 4 liters of still wine, plus 16 liters of beer, per person. This is handy if I visit my relatives in Australia. Mind you, we are only allowed to bring in goods free of duty, the combined value of which does not exceed €430.


Conclusions

As far as buying wine in a liquor store is concerned, Sweden is, in practise, not all that different from anywhere else I have ever been. It is no harder or easier to make purchases, and access is no more restricted — you just have to be organized beforehand. In fact, the main differences are the broader selection (c. 17,500 products per year) and the cheaper prices in the middle and upper quality brackets in Sweden.

On the other hand, at the cheap end, and for impulse purchases, there are some practical limitations — you can't buy strong alcohol in a food shop, for example; but you can't do that in Australia, either (in Australia, the licensed premises must be physically located outside of the supermarket). So, most of the fuss about "the Swedish alcohol monopoly" comes from people who object to government monopolies on principle, rather than because monopolies are costly or inefficient — Systembolaget is neither of these things, in practise.

More to the point, given that liquor sales in so many countries are currently controlled by a few giant supermarket chains (eg. only two main chains in Australia), having a single chain is not all that much different in theory, let along practise.



*1 It may come as a surprise to many Americans that there are still liquor-control states within the post-Prohibition USA. However, Wikipedia notes that about one-quarter of the United States' population lives in the 17 liquor control or government-monopoly states (as of 2016).

*2 This is a pun. "Systembolaget" translates as "The System Company" — it was originally called simply "Systemet" ("The System").

*3 Sweden also has lower taxes on wine than does Britain (see European wine taxes — and what to do about them), although I would not describe either of these taxes as "small".

*4 Compare this with the Pennsylvania Liquor Control Board's 600 stores for 13 million people, and the Liquor Control Board of Ontario's 650 stores and 210 agencies for 14 million people.

*5 This differs significantly from the situation in, for example, Pennsylvania, where: "It is generally unlawful for anyone, other than the PLCB or the holder of a sacramental wine license, an importer’s license, or a direct wine shipper license, to import any liquor (including wine) into Pennsylvania."

Monday, April 30, 2018

Wine yields and latitude

The following post recently appeared on the American Association of Wine Economists' Facebook page. The question posed in the post is an interesting one.


These same data were published by the AAWE back in June 2017, at which time I worked out what the yields are, in fact, related to, since it apparently isn't quality. It turns out to be latitude, with increasing yields as one moves away from the equator, as shown in my first graph. Each point represents one of the French regions, with its latitude plotted horizontally and wine yield vertically.

French wine yields versus latitude

What the line on the graph indicates is that three-quarters of the variation in yield is exponentially related to latitude. The areas named are the main governmental regions of France, so they are very large, and encompass several wine regions — I simply chose the degree of latitude that runs through the region. Obviously, it would be better to have data for each wine region individually, associated with more precise latitudes.

Interestingly, in another AAWE post we do have such data, for the 2016 winegrape yields in the wine regions of Italy. These data are shown in the next graph, where each point represents one of the Italian wine regions. The data reveal four regions (in pink) whose yield is much greater than anywhere else in Italy, or France for that matter — from the top, they are Emilia-Romagna, Veneto, Abruzzo and Puglia.

Italian winegrape yields versus latitude

However, as indicated by the line on the graph, the remainder of the Italian winegrape regions do show the expected relationship between yield and latitude — however, this time only a bit more than a third of the variation in yield is exponentially related to latitude. This difference (35% vs 74%) might be the result of using smaller geographical areas, or it might be related to the use of winegrape yield instead of wine yield (the latter must be less than the former).

Finally, yet another AAWE post shows the average wine yields for selected countries in Europe in 2008. Note that Germany, which is north, has higher yields than France, while Spain and Portugal, which are south, have lower yields than France. So, the trend between yield and latitude is repeated even at the geographical level of country.


If you want a more global perspective on yields by country, then you can check out this other AAWE post: Average vineyard yields 2000-2015.

Conclusion

In western Europe, wine yields are strongly related to latitude, rather than necessarily being related to quality. However, it is well known that "correlation does not equal causation" — latitude is unlikely to be having a direct effect. Whether the proximal effect of latitude is via temperature, or some other factor such as day length or seasonal water availability, is not clear.

Monday, April 23, 2018

Laube versus Suckling — do their scores relate to wine price?

The short answer is: yes, and no!

A few weeks ago I wrote about the wine-quality scores from different professionals, and what their differences might mean for us (Laube versus Suckling — their scores differ, but what does that mean for us?). Another issue of importance to us is whether wine quality scores are related to the prices of the wines — after all, wine could never be good value for money unless higher prices reflect higher quality. So, an obvious follow-up question is whether critics' scores relate to the prices of the wines being tasted.


I have written before about the relationship between wine price and quality (The relationship of wine quality to price), noting that this often follows a simple exponential relationship, as QPR (quality-price ratio) so often does in economics. I have also noted that wine prices often have at least two very distinct price categories (Luxury wines and the relationship of quality to price), with different quality-price relationships. Both of these concepts are relevant to this blog post.

The data I am looking at here is the same as for the prior post concerning James Laube and James Suckling, when they both worked for the Wine Spectator magazine. These data are from the "Cabernet Challenge" of 1996 (see Wine Spectator for September 15, 1996, pp. 32–48), in which the two James tasted 10 California Cabernet blends and 10 Bordeaux red wines from both the 1985 and 1990 vintages. This gives us 40 bottles of wine with which to compare their scores, along with the prices of those same wines, as given in the magazine article.

These data are shown in the first graph, with each point representing a single wine, with its price indicated vertically and the quality score indicated horizontally. Note that each wine appears twice in the graph, once for Laube's score and once for Suckling's (they will thus appear as horizontal pairs, since the pairs have the same price).

Wine price versus quality for all of the wines

The main thing to notice about the data at this stage is that there there are clearly two wines whose prices are disconnected from those of the other wines — Château Margaux 1990 (the pair of points at the upper left) and Château Latour 1990 (the pair at the upper right). Of these two wines, the Château Margaux is a rip-off according to both critics, as its price is completely out of line with their assessment of its quality. Indeed, this wine was singled out in the magazine article as being especially "disappointing".

We can now look at the price-quality relationship of the remaining wines (ie. ignoring these two "luxury" wines), as shown in the next graph.

Wine price versus quality, excluding the most expensive wines

At first glance, there does not seem to be any clear relationship between quality and price — wines with lower quality (at the left) seem to cover a wide price range, as do wines with higher quality (at the right). However, there are actually two factors obscuring the expected relationship here (higher quality for higher price).

The first factor is that James Suckling's scores do show a relationship between quality and price while James Laube's do not. This is why there is a "yes and no" answer to the title question — "yes" for Suckling and "no" for Laube. Formally, the mathematical correlation between wine price and quality for Suckling is 18% while it is only 1% for Laube. [Note: 18% is still not a high value; see below.]

The reason for this difference in the two James is shown in the next graph, where I have drawn ovals around three of Laube's points. These three points are what we call "influential values", in the sense that they strongly affect his QPR correlation — without these three values his correlation would also be 18% (as it is for Suckling).

The main pattern, plus the unusual wines

The two points at the upper left are two Bordeaux wines that Laube scored very low compared to Suckling (Château Mouton-Rothschild 1985, Château Margaux 1985) and the bottom right point is a California wine that Laube scored much higher than did Suckling (Beringer Private Reserve 1990). In all three cases, Laube's score is out of line with the wine price whereas Suckling's score matches the price more closely.

The second reason for a wide spread of points is indicated by the box I have drawn around the pair of points (in the graph) for one of the wines. This wine (Chateau Montelena Napa Valley 1990) is the most obvious bargain among the wines — it has a high score from both critics but is one of the cheapest of the wines. This is a case of quality being high but not price.

Finally, the line drawn on the graph represents the exponential relationship between price and quality for the remaining wines (ie. ignoring Laube's influential values). That is, price does generally increase with quality, with a correlation of 24% (as indicated on the graph). In other words, about one-quarter of the variation in wine price is correlated with wine quality (irrespective of which of the two critics provides the score).

I will leave it to you to think about what the other three-quarters of price variation might be related to!

Conclusion

There are four conclusions from this dataset, which may be true for other data, as well. First, there are luxury wines whose exorbitant price is unrelated to their quality. Second, some critics have scores that are related to price, while others do not. Third, wine bargains can be found. Fourth, only about one-quarter of wine price is related to quality.

I doubt that more recent wine scoring is any different from this (although the prices now are certainly higher!).