Rant on Rank

When I decided to start this blog, I thought a bit about ranking verses rating. Are we actually rating whisky or just ranking whisky among the ones we have tried ourself? And what is the difference? I think of rating as assessing the whisky individually and score it based on its individual merits. Rank a whisky would be to access it in contrast to other whiskies. A ranking can of course be expressed as a number as well? So then you might still say, what is the difference in the end? Can I not just rank the whiskies I tried based on their rating? Yes, you can, but...

I like to draw analogies to sport rankings. If you have two teams; A ranked 1 and B ranked 2, does that mean that  A always beats B? No. Let say you have 3 teams: A,B and C and A beat B, B beat C and C beat A in a round-robin match up. How would you rank the teams? The same questions can be applied to whiskies to some extent. Rating seems to absolute to me. One day I might enjoy Whisky A over Whisky B, the other day the opposite. Sometimes I find whiskies offer different qualities and can’t always imply that since A > B and B > C then A > C. As in sports, ranking changes over time and there are many ways to compute and express ranking, most of them overkill if applied to whisky (and for a blog like this). But I still like the Idea of giving whisky a non absolute rating, that can be updated over time, theoretically, to get the exact ranking, I would have to, every time I try a new whisky, retry all other whiskies I’ve tried before and re-assess their score against the new whisky. Of course practically impossible.

So initially I will try grade my whiskies on a scale from 0-10, with a twist, I will see it as a normal distributed probabilistic score, expressed as r±s , i.e a score that with (my) 95% confidence falls within 2 standard deviations (s) of r.

E.g. Let's say first time I grade a whisky I give it 5±.5 (meaning, with 95% confidence my subjective score falls between 4 and 6), the score could naturally become more uncertain over time, expressed as a larger deviation e.g. 5±1. But If I re-taste the whisky at a later occasion, I can re-asses the score, and hopefully reduce the deviation (the opposite can also happen).

Will I compute it exactly and update it every time I try a new whisky? No of course not…