the Great Divide

[Glenn E.]

There's a lot of talk about how scores have crept up over the years. About how it's easier to get a 90, or a 92, or whatever score you want today than it used to be. Many people feel that ratings should be evenly distributed. Others feel that if it's a 95 it should be rated a 95 whether or not the "quota" of 95s has been used up. Then there those who feel that the ratings system covers every wine ever produced, and as such is "fixed" throughout time. Many others feel that ratings should "shift" with the times so that an "average" score is always average.

Tonight Roy and I were talking about how there are so many good Ports out there, that it just doesn't make sense to drink "bad" Port. (We used that "bad" descriptor loosely... we're really talking about very good Port vs excellent Port.) We used 90 points as the threshold.

So then we got to thinking... what percentage of the Port that you drink do you rate 90 points or higher, vs the percentage that you rate 89 or lower?

I don't have a good feel for it, but I suspect that my 90+ Ports greatly out-number my 89- Ports. Even counting recent tastings where I rated quite a few in the mid to upper 80s, I suspect that overall probably 80% of the Port that I drink or taste is worthy of 90 points or more. I just don't see the point in buying anything less.

How about you?

[Marc J.]

I'd say that probably 50% of the Port I drink falls somewhere between 85-90 points. The other 50% would represent the 90+ point wines. Of course there are some outliers, but I'd say that the vast majority lie within that range. More to the point, I do believe that 90+ scores have become much easier to come by.

[Tom Archer]

As I've said many times before, absolute scoring systems are fundamentally flawed. Relative scoring, where the proportion of players at each grade is fixed, is the only meaningful way of rating wines.

Moreover, so much can happen - both good and bad - between release and maturity of a vintage port, that it beggars belief that scores given 20 years ago or more for newly released VP are still taken seriously.

[Andy Velebil]

...

Moreover, so much can happen - both good and bad - between release and maturity of a vintage port, that it beggars belief that scores given 20 years ago or more for newly released VP are still taken seriously.

So true. I laugh when I'm at a wine store and some employee tells me this Port got rated XX Points and I know that score was from 15+ years ago and is no longer relevant. But numbers sell and it's easier for these employees to spout out numbers than actually learn about the wines they sell.

[Andy Velebil]

As for my break down..I have no idea other than most Port (and wine) I drink are around the 85 point range and higher. I see no need to buy lower tier stuff when for a few bucks more I can get a much better bottle.

[Bradley Bogdan]

I'm on the fence a bit between Glenn and Andy. The lowest tier ports I usually drink are LBV and 10 year Tawnies, as I don't drink port on a daily basis. I suppose if I did drink port on a daily basis, I'd mix in more ruby reserves for cost purposes.

Barring the flawed or truly odd bottles, pretty much everything I drink is worth an 85 or above, and has the potential on a good bottle/day to be 90+ to me.

I've debated with folks many times about score inflation, part of it is certainly the improvement of wine in every region over the last 40 years. I don't see a reason to penalize the average wineries with a curved scoring system because everyone has raised their game.

To me, the point of any evaluation/grading system is to meet the goal of the user. In this case, I use scores as a tool to help me decide whether a wine is worth buying/drinking or not. Normalizing those scores to a curve doesn't help me do that, I don't care if 93 is the new 90 due to recent winemaking improvements, I know that (in theory) that wine will be delicious and of course I want to drink it.

[Glenn E.]

As I've said many times before, absolute scoring systems are fundamentally flawed. Relative scoring, where the proportion of players at each grade is fixed, is the only meaningful way of rating wines.

I have the exact opposite view. Relative scoring is practically useless because it means a wine's score can change dramatically from one session to the next. You're also telling me that a 90-point wine today means something completely different than that same 90 points meant 20 years ago. How is that useful? It tells me nothing.

A fixed ratings system is the only way for the scores to be useful over time - it tells you where every wine falls on an absolute scale, and that number remains constant over time. A 90-point wine today is equally as good as a 90-point wine was 20 years ago. As winemaking improves, you should naturally expect to see higher scores because the wines are getting better.

Under a relative system you cannot see that improvement, and in fact a better wine could end up with a worse score. How does that make sense?

Moreover, so much can happen - both good and bad - between release and maturity of a vintage port, that it beggars belief that scores given 20 years ago or more for newly released VP are still taken seriously.

I completely agree with this - wines change over time, so the date that the rating took place is almost as important as the score and TN text.

[Jim R.]

I am in complete agreement with Glen!

[Tom Archer]

A fixed ratings system is the only way for the scores to be useful over time - it tells you where every wine falls on an absolute scale

Sorry Glenn, but unless you accord demigod status to the various wine critics, it just doesn't work in practice.

One of my favourite mantras to newbies on the fine wine trail is 'never let someone else tell you what is good - be your own judge'

Wine critics can only provide pointers, and, on average, pretty mediocre pointers if the truth be told.

Their absolute scores are very unreliable indicators. However, if they say that 'x' was better than 'y' which was far better than 'z' then you have a personal but relative assessment that is far more informative to the person who has yet to sample the wines in question.

[Glenn E.]

A fixed ratings system is the only way for the scores to be useful over time - it tells you where every wine falls on an absolute scale

Sorry Glenn, but unless you accord demigod status to the various wine critics, it just doesn't work in practice.

With regard to the various critics, it works no better and no worse than a relative system.

What is better: a poor attempt to use a standard system, or a poor attempt to use a system that means something different every time it is used?

One of my favourite mantras to newbies on the fine wine trail is 'never let someone else tell you what is good - be your own judge'
Wine critics can only provide pointers, and, on average, pretty mediocre pointers if the truth be told.

Good advice regardless of the rating system being used, but that begs the question: what does the fact that critics aren't reliable have to do with the system they use?

Their absolute scores are very unreliable indicators. However, if they say that 'x' was better than 'y' which was far better than 'z' then you have a personal but relative assessment that is far more informative to the person who has yet to sample the wines in question.

"X is better than Y which is far better than Z"
is the same as
"X scores 90. Y scores 88. Z scores 80."

The difference is that in the latter case, I can figure out where "W scores 85" fits in the spectrum without needing to compare it directly to Y and Z. You cannot do that with a relative system. Furthermore you're just using abstract numbers - "better" and "far better" instead of a 2-pt spread vs an 8-pt spread.

I also don't buy that there should be an equal number of wines at each relative tier. That's an artificial limitation that assumes there are an equal number of wines spread across the entire spectrum of wines. That simply isn't true, which makes the relative tiers inconsistent.

[Paul Fountain]

I think, that when it comes down to it, I'm looking for a great QPR. That might mean that I drink a nice VP on occasion but it also means that I'll be drinking at least as many well priced ruby reserves and LBVs.
The same is true of both table wines, where I always have a stash of quaffers that are excellent at their pricepoint, and beer as well (as I'm not always in the mood for a double IPA or a 10% stout)

[Al B.]

I generally drink LBV or VP, with the odd bottle of colheita, tawny or ruby reserve thrown in to the mix. In general I will drink things because either:
(a) I know they will be good (which I would define as 88+ points) if the bottle has been well looked after; or
(b) I am curious because it's something new to me

Occasionally I am disappointed under (a), normally because the wine is shut down or hasn't been decanted for long enough - but that's my fault and not the wine's.

Sometimes I am disappointed by (b) (I strongly recommend you avoid Royal Oporto 1983), other times pleasantly surprised (Van Zeller 1987 is delicious).

Most of my drinking is down to the generosity of my fellow port lovers, but if I look at what I pull out of my stocks to share with them or to drink on my own I can see the following pattern in my average scores over the last few years:
2014 - 88.8
2013 - 88.2 (which included 45 shippers of 1963 ports)
2012 - 88.7
2011 - 88.0
2010 - 87.2
2009 - 88.7
2008 - 89.8
2007 - 88.8
2006 - 88.9

The balance of what I drink from my own stocks has stayed pretty constant in terms of age, shipper mix, type of port. I don't see a huge amount of evidence from this quick summary for either inflation in my scoring or for an improvement in the quality of the wines that I am choosing to buy and drink.

On the other hand, if I look across everything I taste in a year - including new release ports and big, organised tastings - I come up with the following:
2014 - 88.8
2013 - 89.4
2012 - 87.8
2011 - 88.1
2010 - 88.1
2009 - 88.7
2008 - 89.8
2007 - 89.9
2006 - 89.9

Again, no real evidence that I'm seeing an improvement in the quality of what I taste or in suffering from score creep.

And if I look at what I've drunk and tasted so far this year, 40% of what I've tasted I scored at 90+, 60% at less than 90 points. That's probably typical for my drinking over a year. My purely emotional view is that I'm seeing fewer bad or poorly made wines these days, partly because they're not being made and perhaps also partly because I know what to avoid. I think there's a much tighter clustering around the mean scores I've given above than there used to be - but I'm not going to sit down and look at how the standard deviation of the scores has changed - there's more port to be drunk!

[Moses Botbol]

I can't speak for what someone else scores a port, and scoring high or low plays no bearing into my decision. Only whether I already had it or by the Producer and/or vintage reputation. I avoid known "issue" bottles like Cockburn 83 or Dow 77 regardless of how epic they are when they're hitting on all cylinders.

Generally, my scores are right in line with you on

[Bradley Bogdan]

In a similar vein, I was thinking last night, as I sipped on some Fonseca Bin 27, that while I "grade" port on a 100 point linear scale, that's not really how I enjoy it. Below my 88ish point threshold, there is a marked drop in my enjoyment of the port for the port's sake (meaning it can be wonderful for other reasons, just not because of the juice itself). Bin 27 is something I usually score in the 86-89 range depending on freshness and such, and this bottle was at the low end. The experience outlined for me that that 88ish point mark is my personal threshold for quality, not just a sweet spot for QPR.


Sent from my iPhone using Tapatalk

[Glenn E.]

The experience outlined for me that that 88ish point mark is my personal threshold for quality, not just a sweet spot for QPR.

It's almost as interesting for me to hear about these thresholds.

The way I score is actually based (more or less) on this concept. My threshold theoretically should be at 85, which is where "very good" starts for me, but over time I've found that it is much more often at "excellent" which starts at 90. Ports that I've scored 85-89 are ones that I like and would be happy to have offered to me, but which I'm not likely to go out and buy for myself. I'd rather get something in the 90s because I've found that they're often the same price.

[Roy Hersh]

Great topic.

Sadly, bottle variation enters into the equation all too often when it comes to older bottlings. As Alex mentioned in a quip on a different dynamic, " ... not the wine's fault."

Rating wine and being held accountable for scores is fraught with peril. It is important to alert readers on occasion, and I do so only in reports on brand new releases, that my score is only the opinion of one man and as Tom mentioned, it would be best for consumers to try the wine for themself (if possible). I don't often have as much issue with pro wine writer's ratings as much as I do with their drinking windows, or lack thereof; But that is a horse of another color.

[Andy Velebil]

One thing that I feel is very important is reviews some years later. As that is really the true test to see if the early reviews were on track or missed the mark.

[Alan McDonald]

So then we got to thinking... what percentage of the Port that you drink do you rate 90 points or higher, vs the percentage that you rate 89 or lower?

Few people have actually answered your question. I have never had the personal need to score wines. I like or I do not like. Some are better than others, but to answer your question, almost all the Port I drink (some Ruby every night with cheese and biscuits, and maybe a tawny with roast almonds, although most often a Moscatel) would be below the threshold of most posters.

That does not bother me. I enjoy what I drink - I simply do not buy those I do not like, and there are too many, and would rather have what I drink every night than have to justify opening a bottle that costs more than I spend in a week. I do have some VPs and open one now and again, but I have ceased buying VP for myself, and earmark those I do buy for my son or grandchildren.

[Roy Hersh]

There are times where I will open Reserve Ruby Ports for an everyday quaffer, or even a LBV Port because I just want to have the flavor of Port but am not looking for an epic experience. I am fine drinking 85-89 point Ports at this time and often really enjoy them regardless of their ratings. Scoring Ports is a great way to reflect and have a benchmark that can be looked back upon to provide a data point of where a Port was at a specific moment, but I don't get all caught up in chasing only Ports that get great ratings. I get the feeling that many of us here have a similar feeling in that regard.

[Al B.]

I am fine drinking 85-89 point Ports at this time and often really enjoy them regardless of their ratings. Scoring Ports is a great way to reflect and have a benchmark that can be looked back upon to provide a data point of where a Port was at a specific moment, but I don't get all caught up in chasing only Ports that get great ratings.

That is exactly how I feel. I enjoy port that I score in the 85-89 point range and will frequently choose to open a bottle which I expect to be of that quality. Occasionally, when sharing a bottle with friends or when I've had a bad week at work, I will open something I expect to be 90+/100. I like variety rather than greatness but, like Allan, see no point in buying or opening things I know I won't enjoy.