Interpreting scores out of 20 and 100

[question]

Question submitted by Jim Gurieff

Continuing on from the 20 and 100 points system discussion, why do we we never see scores of 10 or 50? I guess I could never understand why 10 or 50 did not represent the average or median, and so why does a wine with “something wrong” still score 75? Shouldn’t this suggest that we should start at roughly 15 and 75 respectively which means a 5 and 25 scoring system instead?

[/question]

[answer]

The one consistent aspect of most scoring systems is that they tend to push scores up towards the top end. It’s simply a fallacy that the 100-point system is in fact a system that rates wine with 100 possible levels. For instance, an average wine without much going for it but without much in the way of a serious flaw is likely not to rate below 80 points. Given that only the very greatest wines could possibly score more than 98, the 100-point scale actually only delivers less than twenty gradations between dead boring and dead brilliant. Perhaps it would more accurately be described as a 20-point scale?

But what of the actual 20-point scale itself? The same wine that scores 80 points will rate 14.5 out of 20 according to the conversion scale I use. Given that most people score in units of half marks or whole marks, the equivalent number of gradations for the example I used above is just 11, ie the number of possible scores between 14.5 and 19.5. I leave 20 out of this discussion, since I am simply too optimistic to think I have yet tasted the perfect wine.

In other words, when used conventionally – as indeed they are just most of the time – virtually all drinkable wine must be slotted into just 18 or 19 gradations with the 100-point scale, and a mere 11 for the 20-point scale.

I still score wine using the 20-point scale, but I do it differently. I use the entire decimal point, so I might mark a wine in this scale with scores like 14.7 or 18.3 for instance. This opens up the field of possibilities immensely, giving me a potential 54 gradations of score between 14.5 and 19.8, which still keeps me safely shy of my self-imposed conviction that I am not going to give a wine a perfect score. Yet!

Now, I wouldn’t claim to be 100 statistically perfect here, in that I would not expect to give the same wine precisely the same score in different circumstances of environment, storage, cork quality or whatever, but it does give me room to be more expressive and communicative, which is what I believe the role of the critic to be.

Now, to address your question directly. Given that nobody of any actual influence does it, why would a critic wish to use the 100-point or 20-point system in a way that would directly mean that there was no value in comparing their score to someone else’s. When I was figuring a way to convert my scores from 20 into a score from 100, I was grateful that James Halliday cooperated so willingly with me to ensure that we were both using sufficiently similar logic to ensure that our readers and subscribers could reliably compare our ratings.

While there is absolutely nothing to prevent it, there is no logic in someone deciding to score wine along the lines you mention in your question. Provided, of course, that they do not become the next Robert Parker with enough influence that virtually everyone else ends up following suit.

So, it’s unlikely that you will ever see scores of 10 and 50 out of 100, since nobody would understand them. But, as you then suggest, when interpreting scores out of 20 and 100 respectively, keep in mind that ratings of 12.5 and 70 mean that regardless of their possible ‘pass’ mark, they are barely fit for human consumption!
[/answer] Please login to post comment