The Old Man of Huy wrote recently about whisky ratings, and I got mentioned.
I thought that this was a good time to talk about how I rate whisky.
The basics, first of all - I rate out of 5. And I use half points. Some people don't like that. That's fine - they can use their own rating system! :-)
I use 2.5 as "average", and try to arrange my ratings around that.
0 is technically "no rating". I try not to regard 0 as a usable part of the scale, but more way of recording that I'm not capable of categorising this whisky, and wouldn't want to. It does usually translate into "undrinkable", but might not always do so... If I could only nose the whisky due to medication and someone forced me to give a rating, it would have to be 0 even if I liked the nose!
I want two things from my ratings system. I want it to be easy, and I want it to be useful.
Let's look at easy first.
Easy is why I use half points. I know that there are situations when I will be tempted to use half points. After all, I'm drinking and therefore may not have perfect clarity of mind!
Imagine you have two whiskies in front of you at the same time, and both are deserving of a certain rating. Let's say that rating is eight out of ten. Neither of them is a nine out of ten. But one of them is better than the other. And you want that to be reflected. And now you're tempted to use a half point.
If you do use that half point, you're now rating out of twenty, not ten. So your scale has just been made less useful. Also, what happens when you find a whisky which is better than your 8.5, but not a 9? Do you use quarter points?
Using a rating out of 5 and allowing half points is basically a quick trick I use against my mind, to work around any doubts and prevarications that I may have when assigning a rating. Using a half point between 3 and 4 feels much better than agonising over whether it's a 6 or 7. I sometimes suffer from "kid in a sweetie shop" syndrome, in which the more choice I have the more my decision making process begins to freeze up - so a short scale works better for me!
Scoring out of 5 is also useful. I want to know what I thought of a whisky without any hesitation or second guessing. Let's imagine I'm looking at my shelves, at two bottles. I want to take just one of them to a party, or use it as a gift. Both are old purchases, which I enjoyed but don't have fresh memories of. Should perhaps skip over this bottle that I rated 82 in favour of an 85 that's sitting next to it? Do I really remember why that three point difference exists?
I know myself fairly well - I'll probably feel a smidgen of guilt at picking something less excellent for my friends, over a measly three points. It will delay the choosing, and cause me unnecessary inner turmoil.
Both would be a 4 in my scale, so I don't get "kid in a sweetie shop" and I don't get any guilt or turmoil.
That's useful - far more useful than the appearance of granularity that a wider scale could give me.
Of course, in real life, I can use other factors such as the preferences of those I'll be drinking with to help me choose as well. But a small scale is a more useful scale in practical terms. Larger scales are only useful for intellectual masturbation.
Things cluster, and I'm crap
So I use a scale of 0-5 because it's easy and it's useful to me. But there's also two elephants in the room when it comes to whisky ratings - clustering, and consistency.
Regardless of the scale used, the ratings naturally cluster together because most whiskies are good - if not excellent. If I did a graph of my ratings, I'd see a lot of ratings between the 3 - 4.5 mark. I'd have a rapid drop off down towards the 1 - 2.5 area, and a drop towards 5.
(Because I'm not shy about giving out 5s. I'd don't feel that the maximum should be unobtainable, and I like the fact that I have a few bottles that I have given the maximum rating possible to - despite being able to line them up in order of preference, in a perverse kind of "scale within a scale".)
This clustering is why people like granularity. They're kidding themselves that only using 60 or 70 upwards out of a possible 100 means that they have a useful scale, when they don't. We can partly explain the lack of use of the lower part of the scale because whisky is generally very good these days - everyone is seeking to produce or drink decent whiskies, so only the upper half of any scale will see any great traffic.
Be honest. If you go back and taste the same whisky two months later, will you really give it the exact same score out of 100? If you were tasting the same whisky blind two months apart, would you be willing to put a large amount of money on giving exactly the same score out of 100?
Out of 10?
Out of 5?
Whisky tasting is affected by your company, your mood, the weather outside, what you last ate and how long ago you ate it... The variables are endless. We gloss over them, pretending that we're tasting machines - but we're not.
A friend of mine has a database of whiskies he's tasted, and adds new entries for each tasting of each whisky. So he can calculate an average. That may be going too far, but it has at least shown that his ratings do vary. And I think everyone should do something similar as an experiment, just to see how their views can differ... I've been surprised when comparing my previous ratings with the whisky I have in my hand!
No rating system can effectively deal with the fact that things cluster, or that humans aren't actually whisky tasting machines.
So I just went with what was best for me. You should too.