Red Burgundy Research. Part 2.1. The Best Grand Crus

Dear all,

I recently published Part 1 of the Red Burgundy Research. The first part was about the best villages producing Pinot Noir in Burgundy.

In the second part of the research I focused on the Grand Crus and how some of the Premier Crus relate to them.⁣⁣⁣

Follow the link to read the article - Red Burgundy Research. Part 2.1 The Best Grand Crus

I am very grateful to many of you who have contributed to the discussion of the first part. I tried my best to reflect on that in my second article and also share some observations.
⁣⁣⁣
I invite you to a discussion. Which are your favourite vineyards of red Burgundy?⁣⁣
 
Hi Stas impressive work. One thing I couldn’t find was an example of the method you’re using for this. What is your formula to calculate these rankings? Without that I’ve really no idea whether what you say makes any sense. Thanks.
 
Last edited:
Hi Stas impressive work. One thing I couldn’t find was an example of the method you’re using for this. What is your formula to calculate these rankings? Without that I’ve really no idea whether what you say makes any sense. Thanks.

Hi Gareth!
Thank you for acknowledging my efforts here. Methodology is based on analysing data from reviews of wine critics, wine journalists and other professionals.
 
Hi Gareth!
Thank you for acknowledging my efforts here. Methodology is based on analysing data from reviews of wine critics, wine journalists and other professionals.
Sorry I realise that I just wonder what your actual detailed method is? Do you take average ratings for each vineyard from every producer in every vintage and convert that average to be expressed as a %? I can see how you might do that but your results make no sense as the results are too high. Likewise how did you produce the actual numbers for your village ranking which are expressed differently (are these additions or ranked out of some theoretical total)? It’s really difficult to interpret what you are suggesting without this basic information. Is a key component your own weighting/scoring? Or is it just a formula on existing scoring metrics?

could you give an example calculation for a grand cru and a village?
 
Sorry I realise that I just wonder what your actual detailed method is? Do you take average ratings for each vineyard from every producer in every vintage and convert that average to be expressed as a %? I can see how you might do that but your results make no sense as the results are too high. Likewise how did you produce the actual numbers for your village ranking which are expressed differently (are these additions or ranked out of some theoretical total)? It’s really difficult to interpret what you are suggesting without this basic information. Is a key component your own weighting/scoring? Or is it just a formula on existing scoring metrics?

could you give an example calculation for a grand cru and a village?

Thank you Gareth for your questions. I will try my best to explain.

First of all, I suspect that the numerical rating is what brings the confusion. The appellation which scored the highest, I gave 100 points. So you see Romanee-Conti having 100. All other appellations received their respective ratings according to their score and adjusted to 100.

Let me give you a bit more specifics below.

Scores are collated from a wide range of critics, wine journalist and other professionals. Then they are all adjusted for the 100-point scale as different critics use different scales. In order to rank the appellations I use the weighted average across all vintages for each specific wine. Wines coming from one appellation are sorted together to identify the top ones. Comparing top lists (top10, top20, etc.) from each appellation with others gives the rating of those appellations. For example, top10 wines from Richebourg get higher weighted average scores than those of Clos de la Roche. Similar to village appellations. The only difference that village appellations have many climats, including communal and 1er Cru level. So in that case we have much more data to process.

Hope that now makes more sense. Let me know if you have any questions to that.
 
If you take the top 10 rather than the mean score of all wines, doesn't that favour larger crus that have more wines to choose from? That could explain why Echezaux scored higher than GE, for example (which I saw someone moaning about on FB the other day).
 
If you take the top 10 rather than the mean score of all wines, doesn't that favour larger crus that have more wines to choose from? That could explain why Echezaux scored higher than GE, for example (which I saw someone moaning about on FB the other day).

Thank you Nick for raising this question.
Top 10 was just for the purpose of providing the example. It is bit more complicated than that.
In case with monopoles, you have only 1 wine and that's it. All you can do is come up with weighted average score across the vintages.
As for the appellations where you have many different wines/labels the approach is to compare topX wines rather than average or mean or median. The idea is to see what the terroir is capable of in the hands of the best winemakers across the vintages rather than getting the average temperature of the patients in the hospital, which doesn't really tell that much. That is the chart of terroirs and in case of those vineyards where you have a huge number of low profile negociant labels, etc. you dilute its value.

My rule of thumb is following this strict order:
1. Winemaker
2. Terroir
3. Vintage
 
That's a fair point of view, although I suppose you could then say it's also a reflection of which crus are lucky enough to have the most competent producer(s) rather than the terroir per se. That information does have value given you (and most people) would say producer is the most important factor...
 
Thank you Gareth for your questions. I will try my best to explain.

First of all, I suspect that the numerical rating is what brings the confusion. The appellation which scored the highest, I gave 100 points. So you see Romanee-Conti having 100. All other appellations received their respective ratings according to their score and adjusted to 100.

Let me give you a bit more specifics below.

Scores are collated from a wide range of critics, wine journalist and other professionals. Then they are all adjusted for the 100-point scale as different critics use different scales. In order to rank the appellations I use the weighted average across all vintages for each specific wine. Wines coming from one appellation are sorted together to identify the top ones. Comparing top lists (top10, top20, etc.) from each appellation with others gives the rating of those appellations. For example, top10 wines from Richebourg get higher weighted average scores than those of Clos de la Roche. Similar to village appellations. The only difference that village appellations have many climats, including communal and 1er Cru level. So in that case we have much more data to process.

Hope that now makes more sense. Let me know if you have any questions to that.
As I said I can understand the method and that seems very sensible but it simply wouldn't spill out the results you have in your table without another step. Take one example. Domaine de la Romanee-Conti Romanee-Conti Grand Cru. I inputted all the data from wine-searcher pro scores from the 2016 to 2000 vintages (a smaller sample but illustrative), this totalled 111 tasting scores. For those marked out of 20 I converted them to 100 point scale and then I averaged the lot. From that I got an average of 96.941.

So I just don't understand how you have gotten to your final number for that vineyard. You must have weighted them differently once you have averaged out the scores as I don't see that the vineyard example I use could ever achieve anything near a 100 point average and I would agree with you that this is the top vineyard. This would be even truer of Chambertin which has many poor wines year in year out, especially if you go back a few years.

Might you please show your workings for Domaine de la Romanee-Conti Romanee-Conti Grand Cru so we might see how you are actually calculating things?

Please understand I do think your ordering of vineyards has merit and meaning I just think the scores would be lower and don't understand how they are so high.
 
As I said I can understand the method and that seems very sensible but it simply wouldn't spill out the results you have in your table without another step. Take one example. Domaine de la Romanee-Conti Romanee-Conti Grand Cru. I inputted all the data from wine-searcher pro scores from the 2016 to 2000 vintages (a smaller sample but illustrative), this totalled 111 tasting scores. For those marked out of 20 I converted them to 100 point scale and then I averaged the lot. From that I got an average of 96.941.

So I just don't understand how you have gotten to your final number for that vineyard. You must have weighted them differently once you have averaged out the scores as I don't see that the vineyard example I use could ever achieve anything near a 100 point average and I would agree with you that this is the top vineyard. This would be even truer of Chambertin which has many poor wines year in year out, especially if you go back a few years.

Might you please show your workings for Domaine de la Romanee-Conti Romanee-Conti Grand Cru so we might see how you are actually calculating things?

Please understand I do think your ordering of vineyards has merit and meaning I just think the scores would be lower and don't understand how they are so high.

Dear Gareth!

You are absolutely right. That was pretty much the route I was taking.
So if I take your numbers as the example. You get 96.941 points for Romanee-Conti Grand Cru. That was the highest scoring vineyard in my research, so I assigned it 100 points, just for the simplicity. Let the champion be 100. That means that I used the multiplier of 1.0316 to get from your 96.941
I did similar calculations for other vineyards. Let's say for La Tache I had 95.938 Applying the same multiplicator of 1.0316 that gives 98.97 in my chart

Think of Romanee-Conti Grand Cru as 100%. So La Tache would be 98.97%

I believe you considered 100 for Romanee-Conti Grand Cru as 100 points on certain scale. I simply adjusted my not very even number I got for Romanee-Conti Grand Cru to 100 and than used the same multiplicator for other vineyard scores. These are the numbers you see in the chart.
 
That's a fair point of view, although I suppose you could then say it's also a reflection of which crus are lucky enough to have the most competent producer(s) rather than the terroir per se. That information does have value given you (and most people) would say producer is the most important factor...

Dear Nick!

Bingo!!! Absolutely spot on!!!!

I am so glad it was not me who pointed that out. Thank you for being so observant, Nick! That is exactly the discussion I was waiting for.
I didn't want to nudge in that direction by telling others my conclusions at this point. I will definitely share that in my afterword after I am done with publishing the whole study.

But you are right. That was among the biggest findings of my study. If I put it in simple words, some vineyards are blessed with star winemakers owning land there and producing wines from the vines growing there. That is by far not the only discovery the numbers revealed, but definitely an important one.

If you look back to Part 1 of my research, you will notice Fixin being at the very bottom of the chart. Fixin is the appellation with long history and a number of 1er Crus. But somehow it scores lower than Marsannay which has zero 1er Crus at this point of time. Marsannay is just blessed with some winemakers like Domaine Sylvain Pataille, Joseph & Philippe Roty and Bruno Clair to name a few, who revolutionized it by making outstanding wines from this vineyards.

Also take Domaine Leroy which is the outstanding producer making among the best wines at each terroir they own. There are many examples like that.

Let me share a few examples from Champagne, another region I adore. Think of Cedric Bouchard (Roses de Jeanne), Olivier Collin (Ulysse Collin) and Alexandre Chartogne (Chartogne-Taillet). Those growers actually made their terroirs famous just because they are making wines there. Those places were noname, no status, nothing. Now they are admired all over the world.
Not to forget of Anselme Selosse. Of course he got his vineyards in famous villages, most of the Grand Cru status. But he again revolutionized the winemaking there. As the result, his wines are topping the most of champagne charts.

If I go back to my research, Clos de Tart might be a very good example as well. Being the producer of just singe vineyard, they become the specialists of this terroir. Striving for the best possible quality, thoroughly studying their terroir they developed their own approach by vinifying separately different parts of the vineyard according to soil types. Blending the best of the best. Converting whole vineyard to biodynamic.

Think of Vogue in Bonnes-Mares who also vinify terre rouge and terre blanche wines separately to get the optimal by blending the two. There are many other great examples out there. There are even more bad examples where winemakers unfortunately waste the terroir they are gifted with.

So those rankings that my chart shows give lots of food for thought. Why certain vineyards are ranked higher or lower than expected.
 
Stas,

I read your report. I'm fine with identifying preferences, and making favorite lists is a pastime in many hobbies. But as a statistician, my problem is with your numbers. To put it simply, wine scores are garbage. And if you are basing your ratings on these scores, then garbage-in, garbage out.

Why is wine scoring garbage?
1) were the scores given wines given in blind tastings? If not, they are biased, which corrupts the stats.
2) are the scores reproducible? In general no, critic scores vary from bottle to bottle. Which leads to...
3) Is the 100 point scale overly precise? Yes, and this false precision makes it look like science rather than pseudo-science. It’s part of the con.
4) Is there consensus on what qualities make wine great, and therefore can be used as scoring criteria? Hell no, especially in Burgundy. Parker and his international style came close to converting the world to his idea of great wine, all power and extract. Where was his greatest failure and retreat? Burgundy, where character trumps power. How do you put character on a 100 point scale? Oak usage and stem usage are two choices which could impact score, but who is to say which decisions are better? If there was an objective truth, all the wine making would converge on that style. Happily, it doesn’t.
5) I could go on. But no, that’s enough.

Go to one of the Burgundy tastings frequently organized on this board. While folks certainly will express preferences, they spend more time making observations. And I’ve never heard anyone score a wine, at least not without irony.

Personally, I would argue that Musigny belongs in your Supreme category. The fact that it only has a weighting of 97.63 (talk about false precision!), won’t convince me otherwise. I’ve got 50 reasons, but a score ain’t one.

If you consider this list of 31 vineyards to be the best, I’m cool with that. If you want to follow Lavalle, Morris, and others, and create a best-of-the-best category, that seems reasonable. But why would you choose to break the best into four groups? Do you have a statistical reason for it? Or do you just think we need to go to a regional<village<premier<third<second<first<supreme?

It’s funny, I don’t hate your final list, I just have a few minor quibbles. But I am biased, I am shaped by the same group-think we all are, including the critics. So of course, critics perpetuate the group think in their scoring, and you perpetuate it in your summary. It makes me wonder, perhaps it’s all a very long con, started by some monks a thousand years ago. (Yes, I’m kidding, mostly.)

I’d happily drink any of the 31 wines with you. But I’d rather talk about the wine's qualities rather than its scores. That is the soul of burgundy.
 
Hi Gareth!
Thank you for acknowledging my efforts here. Methodology is based on analysing data from reviews of wine critics, wine journalists and other professionals.

Are there any professional wine **critics** when it comes to Burgundy? I thought that wasn't possible any more ...
As one very well known wine "critic" told me about a recent vintage, "If I'd written what I really thought about DRCs as they tasted from barrel, Aubert wouldn't have let me go back".
 
Stas,

I read your report. I'm fine with identifying preferences, and making favorite lists is a pastime in many hobbies. But as a statistician, my problem is with your numbers. To put it simply, wine scores are garbage. And if you are basing your ratings on these scores, then garbage-in, garbage out.

Why is wine scoring garbage?
1) were the scores given wines given in blind tastings? If not, they are biased, which corrupts the stats.
2) are the scores reproducible? In general no, critic scores vary from bottle to bottle. Which leads to...
3) Is the 100 point scale overly precise? Yes, and this false precision makes it look like science rather than pseudo-science. It’s part of the con.
4) Is there consensus on what qualities make wine great, and therefore can be used as scoring criteria? Hell no, especially in Burgundy. Parker and his international style came close to converting the world to his idea of great wine, all power and extract. Where was his greatest failure and retreat? Burgundy, where character trumps power. How do you put character on a 100 point scale? Oak usage and stem usage are two choices which could impact score, but who is to say which decisions are better? If there was an objective truth, all the wine making would converge on that style. Happily, it doesn’t.
5) I could go on. But no, that’s enough.

Go to one of the Burgundy tastings frequently organized on this board. While folks certainly will express preferences, they spend more time making observations. And I’ve never heard anyone score a wine, at least not without irony.

Personally, I would argue that Musigny belongs in your Supreme category. The fact that it only has a weighting of 97.63 (talk about false precision!), won’t convince me otherwise. I’ve got 50 reasons, but a score ain’t one.

If you consider this list of 31 vineyards to be the best, I’m cool with that. If you want to follow Lavalle, Morris, and others, and create a best-of-the-best category, that seems reasonable. But why would you choose to break the best into four groups? Do you have a statistical reason for it? Or do you just think we need to go to a regional<village<premier<third<second<first<supreme?

It’s funny, I don’t hate your final list, I just have a few minor quibbles. But I am biased, I am shaped by the same group-think we all are, including the critics. So of course, critics perpetuate the group think in their scoring, and you perpetuate it in your summary. It makes me wonder, perhaps it’s all a very long con, started by some monks a thousand years ago. (Yes, I’m kidding, mostly.)

I’d happily drink any of the 31 wines with you. But I’d rather talk about the wine's qualities rather than its scores. That is the soul of burgundy.

Dear Brady,

First of all, thank you so much for reading my text. I really appreciate your time investment.

I see your point. I did address it in my text as well. I came across similar point of view before. The denial of any numeric score to be applied to the wines of Burgundy. Well, there is the opposite point of view as well. And I am personally fine with both. I enjoy Burgundy wine and observe the nuances, complexity of the aromas and flavours, its length and trying to assess its potential. I simply sit back and enjoy every sip of it. On the other hand I find it very natural to many humans to put things in some structure and order. If one feels like scoring the wine as a way to compliment the expression of ones emotions induced by the wine, I am fine with that as well. Moreover, many great experts who I learnt Burgundy from use numerical scores. And they did help me to structure my thinking and helped my learning the region. I am talking now about Coates, Meadows, Robinson and many others. I might agree or disagree with certain scores, but I do accept them as one of the means of expressing the impression. Such scores, all of them are subjective by nature. No doubts. However, the amount of data that has been created and collected over decades is immense.

It is now very difficult for me to deny the fact that all these scores have no impact at all. They actually do. They create a great deal of an impact. They set the trends. I am pretty confident that probably all of us are biased with those scores in some way. Those of us who have own great experience with Burgundy may have own strong opinion that can be inline with trends or not. That is the diversity of opinions. And that is even better. That creates new thinking, sets new trends, etc.

The idea of my research was to look at the data and see what comes up. That is not the dogma. That is not my personal opinion. That is just an attempt to reflect on current trends and see beyond that.

And one of the findings that has been already confirmed by other readers of my research, that some of the terroirs are blessed by very talented winemakers owning plots there and making wine.

So when discussing the terroir, what exactly do we mean? Are we talking about its potential, its nobility or breed? Or we are talking about the wines made of the grapes coming from such terroir? There is a very finite amount of wines available from each terroir. New generations of winemakers come, new landlords appear, new winemakers emerge, climate changes, farming and winemaking techniques also change. But Richebourg remains Richebourg. What about the wines of Richebourg?

What happened when Henri Jayer passed away. What will happen once Madame Lalou era is over? What is happening to terroir such as Criots-Bâtard-Montrachet when Caroline Morey starts making wine there? How Corton-Charlemagne will change once DRC release their wines?
Have a look at Marsannay appellation. This is where lots of happening right now.

There are many other observations that I have playing with all these scores.

That is the kind of discussion I would love to have. My study is just the invitation.

But one can blame it, attack it and suggest that I shut the f*ck up as scores are bullsh*t. Full stop ))
 
We all enjoy our wine in different ways but I genuinely do not give a moment's thought to scores.

I trust your self awareness, however it might even be unconscious.

The way we navigate when choosing wines we don't know. You might rely on someone's opinion or recommendation. We read through the wine tasting notes. Some of them have or have no numeric scores. Use cellartracker or alike.
It really doesn't matter. We might think we are immune to scores, but it can be asymptomatic ))
 
Stas,
interesting effort. But if Chambertin as a whole is classified & rated higher as Musigny I would suggest you try a Chambertin from Bertagna against a Musigny from Vogue. After that you'll see the problems of your list and classification.
Cheers
Rainer
 
Stas,
interesting effort. But if Chambertin as a whole is classified & rated higher as Musigny I would suggest you try a Chambertin from Bertagna against a Musigny from Vogue. After that you'll see the problems of your list and classification.
Cheers
Rainer

Dear Rainer!

Thank you for your comment. You are making a very valid point here. Let me clarify to eliminate any possible confusion.
Consider this chart as what each vineyard (terroir) is capable of based on the wines available in the market. That brings me back to my rule of thumb stated in Part 1 of my study. Choose the winemaker over terroir over vintage.
So if we take a handful of the best Chambertins and compare them with a handful of the best Musigny. I am talking Leroy, Rousseau, Perrot-Minot, Dugat-Py, Denis Mortet, Olivier Bernstein, Trapet, Dujac, Roumier, Mugnier, Vougeraie, Vogue.
Why comparing the worst with the best? What are you trying to achieve with it?
 
Dear Rainer!

Thank you for your comment. You are making a very valid point here. Let me clarify to eliminate any possible confusion.
Consider this chart as what each vineyard (terroir) is capable of based on the wines available in the market. That brings me back to my rule of thumb stated in Part 1 of my study. Choose the winemaker over terroir over vintage.
So if we take a handful of the best Chambertins and compare them with a handful of the best Musigny. I am talking Leroy, Rousseau, Perrot-Minot, Dugat-Py, Denis Mortet, Olivier Bernstein, Trapet, Dujac, Roumier, Mugnier, Vougeraie, Vogue.
Why comparing the worst with the best? What are you trying to achieve with it?

It doesn’t help that your list omits the best producer of Musigny.
 
Is Chambertin at its best 'better' than Musigny at its best? if one thinks about it for a moment it's not so much a difficult question to answer as a pointless one-it is a choice that few of us have the opportunity or even desire to make at all frequently, and who would not be thrilled by either?
 
Is Chambertin at its best 'better' than Musigny at its best? if one thinks about it for a moment it's not so much a difficult question to answer as a pointless one-it is a choice that few of us have the opportunity or even desire to make at all frequently, and who would not be thrilled by either?

I will drink wines of all 31 vineyards from the best winemakers ))) Not a moment of hesitation.

Actually, I am inviting to explore Burgundy more and share the opinions. My study just reflects the current state. Yes, it seems like wine professionals do treat top Chambertins bit more than they do Musigny. Does it say the one is good the other is not? Not at all. Maybe it tells something about the producers making those wines. It might tell something about the stylistic preferences. It can also give an insight to the winemaking and viticulture approach of the two. I am more interested in those questions and the following answers.

But most of you stubbornly defend your preference and the denial of any measure of Burgundy.

If you allow yourself for a moment to accept the hierarchy the way the study shows, what does it tell you? What could be the underlying factors that influence that? I find it by far more interesting discussion than attacking the numbers )))
 
Top