It seems like I am the Paul Konerko of Wine Blogging. Or maybe I am the Benard King, the Kenny Anderson, or perhaps even the Bernard Thevenet.
Who are all these people and what do they have to do with wine?
Well, they are all sports figures and as far as I know, they have absolutely nothing to do with wine. But they all have something in common. Paul Konerko was a first baseman, mostly for the White Sox, and is 42nd on the all-time home run list. Bernard King played 14 seasons in the NBA and is currently 42nd on the all-time scoring list. Kenny Anderson, the long-time Cincinnati Bengal quarterback, is 42nd on the NFL career touchdown list.
Then there is Bernard Thevenet, the popular French cyclist turned broadcaster, who is listed by the Cycling Hall of Fame as the 42nd greatest cyclist of all-time. It is not entirely clear how the Cycling Hall of Fame derived the list, but they no doubt attributed points and weights to various races, thus coming up with the final “score.”
That brings us to yours truly.
A few days ago, Exel Wines, an online wine shop in Perth, Scotland, put out a list of the “Top 100 Most Influential Wine Blogs of 2015” and there was The Drunken Cyclist at #42. On the website, Exel Wines claims that “never before has there been such a statistical measure of influence in the wine blogging community as what we’ve put together over the past several months.”
Being the math geek that I am, when I read the word “statistical” I immediately become both excited and wary. Excited because, well, I am a math geek and wary because as a math geek I know that people often misuse statistics to prove (or disprove) a variety of things.
Since I have been blogging, there have been a few such rankings that have emerged and each one of them had their positives and negatives, but most have been fairly easy to understand, so I was anxious to see what Exel had done that was different and “statistical.”
So I read on: “the rankings … include a detailed breakdown of the ranking algorithm and methodologies.”
Now I was really excited! Not only was there a ranking, but they were going to share the algorithm and the methodologies. As many of you know, I went through the torture of writing a dissertation, which as a whole is pretty worthless, but I did enjoy the statistical aspect of that work. So much so that I now use it pretty much everyday in my “real job.”
I ignored the Exel rankings at first and just scrolled down to find the “detailed breakdown of the ranking algorithm and methodologies.” For you non-math geek types (of which I am sure that there are a few), in layman’s terms: “in mathematics and computer science, an algorithm is a self-contained step-by-step set of operations to be performed” (Wikipedia).
In short, an algorithm is a formula or a procedure used to solve a problem. They can be relatively complex, or fairly simple, but they all have a series of steps that need to be followed, a recipe, essentially.
The point? While Exel lists the different components (or ingredients) that went into their algorithm, they do not provide any further detail as to the calculations involved. Imagine if you wanted to bake some bread and I told you simply: “Get some flour, yeast, water, and salt, mix it together, and put it in an oven.”
I am afraid that is the extent of the “detailed breakdown of the ranking algorithm and methodologies” for this ranking by Exel Wines and, as a result, the extent of my math-geek excitement.
Generally speaking, in the research field, you instantly become dubious of any study that does not include a detailed explanation of the methodology, since a key to any good study is its replicability (the ability for others to go in and examine the work).
I started looking at some of the data and, well, there is a pretty glaring warning sign right at the top. While there is a gradual decrease in the “Site Score” from #2 through #100, the top rated site, Madeline Puckette’s wonderful Wine Folly, has a Site Score that is a whopping eleven times higher than the second rated blog.
So I did what every good quantitative researcher would do. I plotted the data. Here are the scores for blogs 2 through 20:
So far so good.
Then I included the #1 blog, WineFolly.
I dug a little deeper and compared the top blog, Wine Folly, with the #5 blog, Jon Thorsen’s Reverse Wine Snob. According to Exel’s figures, Jon has 12 times more Twitter followers, half as many Facebook followers, twice as many Google Plus followers, but has a score that is one fourteenth of Wine Folly’s score?
Sure, there are some other metrics that they cite that might play a role, but there is no indication of either what they mean or how they were weighted. I thought about contacting Exel to see if they would send me the actual algorithm, but, well, I already have enough enormous time sucks on my hands.
Don’t get me wrong, I am sure that it took considerable time and a bunch of work to put this together. I am also flattered to be one of the Top 100, but I would advise taking this with a grain of salt, as my grandmother used to say.
But hey, for now? I have no problem being the Bernard Thevenet of Wine Bloggers.