Vintages are like children: all different, all hard to predict and all interesting. When a child is born you can only guess at their future personality. What will they be like in 10 years time? How will they behave when they reach adulthood? Each child is different and they change over time to show different characteristics -- good and bad. But who can genuinely know what will be interesting about a one year old when they reach maturity? Nobody except the wine critics.
I have been tasting in Burgundy since 1992 and the quality producers are bemused at the way the media and wine merchants rush to characterise a vintage after tasting samples from a few producers' barrels and getting weather reports for the Cote d'Or. The reason? People buy vintages in the same way that they buy brands. 2005 is a heavily branded vintage, 2004 is not, neither is 2006. Equally 2003 was heavily branded by the media due to its unique characteristics. But 2002 and 2001 are forgotten and 2000 erased from memory.
So what drives media statements about vintages? They typically get alerted to a vintage by the buzz about Bordeaux wines which vary in price far more than Burgundy does on a vintage by vintage basis. For example, 2005 was slightly too warm for it to be a great white Burgundy vintage but the usually-reliable Guide Hachette scores 2005 as 18/20 and 2004 as 15/20. Even the Burgundy producers think this is wrong! From time to time the big Burgundy negociant like Faively, Champy, Louis Latour, Louis Jadot, and Domain Drouhin need to push Burgundy wines in the media. This is completely appropriate but it does not mean that the vintage they push is indeed the best. It is simply the one that the media will take interest in.
The media also focus on what the mass market will find interesting because they serve the mass market not the connoisseurs (like us!). For example, 2003 was broadcasted as an incredible vintage with masses of super-ripe fruit and was classified by the Guide Hachette - the most influential wine guide in France - as 17/20 for Burgundy reds and 18/20 for whites. But these wines, even the grands crus like Echezeaux, are already hitting maturity. Because over-ripe fruit dominates the structure of the wines and there is so little acidity they simply will not age. For from being a stellar vintage it is a bizarre short-lived vintage with limited complexity and little elegance. Rather like an attractive and voluptuous peasant girl at the aristos ball. Yet the Guide Hachette scores the 2005 whites far above the classic 2004 vintage.
The alert connoisseur also has to be careful about vintages and prices. Fine Burgundy production is dominated by small family owned domaines that need to manage their income over multiple vintages. When they have a small harvest like 2003 when most domaines produced 40% or 50% less than usual they often have to put up prices simply to stay in business. But this does not mean that high prices equate with high quality vintages (as in Bordeaux). In Burgundy, high prices equate with low yields. One therefore needs to avoid infering that the highest priced vintage (2003) is the best because it isn't. Normal or high yielding vintages such as 2005 often produce excellent wines and due to the high yields the price stays low. Alors attention mon ami!
Returning to my opening remark: like children wines evolve over time and develop a personality that is hard to predict. 2000 reds are a great example - the vintage was panned by the media for being too light. But many of the 2000 1er crus and grand crus have now developed mysterious and complex bouquets which are the hallmark of good pinot. A contra example is 1996 which many believed would be an historic year due to the very high acidity. But surprisingly most wines, including the whites, didn't integrate the acidity on the palate, resulting in slightly unbalanced but long-lived wines. This result certainly wrong-footed many of the wine makers who had hung onto their on 1996s. Because you can't have two great vintages in a row 2006 will be over-shadowed by 2005 but the Cote de Beaune producers consider their 2006 whites to be far superior to 2005 and even 2004.
In conclusion, after 15 years of tasting on the Cote d'Or and sampling my private collection over time my advice is to buy every single vintage from good producers you know. Who can argue with that? Test producers you are unsure about by buying their wines from poor vintages. The old adage that any vigneron makes good wine in a good vintage but only the capable and quality conscious make good wine in a difficult vintage holds true. For reds buy from the petits villages that are further north (Marsannay, Fixin) or higher in altitude (Pernand, Saint-Romain, Saint-Aubin) in riper years like 2003 and 2005. Don't forget that chardonnays shine in slightly cooler conditions than the pinots so while 2005 is a cracker for pinots, 2004 is better for the chardonnays and 2003 was overall pretty awful because the whites were cooked. The Burgundians appreciate loyalty so by buying a small amount every year you become a "client fidele" which gives you access to a few bottles of Richebourg 2005!