I love the internet. Speaking as an old man who remembers The Dark Times, I can say that never has so much information been so readily accessible. In the days before free and easy information we would just argue over stuff. Was it Kurt Russell or Patrick Swayze in Road House? Which movie was better? Rocky I or Rocky II? Did Mikey really die from eating pop rocks and drinking soda at the same time? Without having any means of answering these questions (other than going to a public library – yea, right. Good luck with that) there was no way to resolve these questions other than to argue. I mean, just imagine trying to play Six Degrees of Kevin Bacon without access to the internet? It would be impossible. But today, most arguments are avoided completely by simply googling the answer.
But in addition to answering trivia questions, the internet also exists for us to express our opinions. Nowhere are our opinions more important than the internet, where our well-thought out and reasoned opinions are joined by thousands more to create a large caterwauling mess. Which brings me to the topic of this blog post. In addition to loving beer, I also enjoy going to some beer websites and reading reviews about beers that others have tried. There are two great websites for this : Ratebeer.com, and Beeradvocate.com. Both of these websites allow people to rate the beers that they have tried, as well as write reviews about them. I tend to favor BeerAdvocate.com mostly because it also provides information about the styles of beers that are being rated. Both websites have cool features that allow you to sort the beers based upon the ratings. At Beeradvocate.com, they rank their beers (on a 1 to 5 scale) based upon a “WR (weighted rank) [which is] a Bayesian estimate that pulls data from millions of user reviews and normalizes scores based on the number of reviews for each beer or place. The WR represents the beer or place’s score against all others.”. As a statistics aficionado, I had to stop for a moment an appreciate the use of Bayesian estimates, which tend to pull the a ranking closer to a mean value, especially when the reviews are few in number. But their statement also generated a number of follow-up questions, such as whether the Bayesian estimate was pulling the rating toward the grand mean or the style mean? Did they treat raters as being nested within beer? What about the multiple rater problem??? Luckily the beer aficionado in me told the stats person in me to STFU.
So I was browsing Beeradvocate.com and I wondered – what beer styles are rated the highest? Sadly, this answer was not readily available from the pull-down menus (not the first time I’ve been disappointed by a pull-down menu). So I decided to get this information manually. I created refreshable web queries in Excel for each of 84 beer styles listed and populated an Excel spreadsheet with the top 50 beers of each style (along with their ratings). Then I graphed the results. What did I learn (other than apparently I have too much free time)? I learned that the most highly rated beers are angry beers that are high in alcohol (see graph). The bar graph contains all the beers that had an average score of 4 or above. Looking at the graph, it seems like there are 4 styles that stand above the rest of the best – Double or Imperial Stouts, Imperial Russian Stouts, Double or Imperial IPAs, and American Wild Ales. The stouts and IPA’s at the top of the list didn’t surprise me. But the American Wild Ales? These are beers that have an additional strain of “wild yeast” or bacteria added to the wort that can leave a sometimes earthy, sometimes sour taste to the beer. The only American Wild Ale I have seen in these parts of the country are from Ommegang Brewery in Cooperstown NY. Further research seems to indicate that most of the best American Wild Ales are brewed on the West Coast, and they don’t seem to make it this far east. That’s too bad – I would love to give these Wild Ales a good tasting. Another surprise in the top 10 were the Saisons. These beers also have an earthy/sour flavor to them that I didn’t think many people would find pleasing – or rate that highly. Overall, the beer styles that made the top of the list are typically high in alcohol and complex in flavor. They have a lot going on.
On the other end of the continuum, I also graphed out the least favorite styles. Please forgive me, but the vertical axes of the two graphs are not on the same scale. Do note, however, that the top of the scale for the worst beer styles (4.0) is the lowest average rating from the graph above. It’s pretty clear that the worst of the worst are the beers that most Americans drink. Light beers, American Adjunct Lagers (fancy name for the buds and millers of the world) and Malt Liquor. Colt 45 does not work everytime with people who rate beer on the internet. Two styles that made the list that I found interesting were the American Wheat beers and the American Dark Wheat beers. I’m going to take a guess as to why they are down here. Compared to the German Wheats, the American Wheats are too highly hopped. Wheat beers have a certain creaminess to them, and that gets ruined with the addition of hops. Much in the same way no one wants bitter ice cream, no one wants to taste bitter when they are expecting creamy. Another common theme in the bottom list is that almost all of these beers are fairly low in alcohol content and are fairly simple in flavor. Apparently people who love beer love to get buzzed fast. Who knew?
So if you’re interested in knowing more about beer or styles of beer, I urge you to check out beeradvocate.com or ratebeer.com. They really have a wealth of information about craft beers. It can even make going into a craft beer store more fun. Check out some different styles of beer and see which ones you like the best.