Welcome to BeerGraphs!

Eno Sarris, June 07, 2013

Baseball and beer seem hopelessly intertwined. My old joke is that the venn diagram for the two is a perfect circle.

Of course that's not true. But there is something so correct about sitting down to a long, cerebral summer game with a pint of suds. Baseball takes a while and is set up for mellow conversation about the intricacies of style and strategy -- beer takes a while to drink and is set up for mellow conversations about the intricacies of style and brewing. The more you learn about each, the more fascinating the two arts can become. And no matter how much you know, there will always be a play or a beer that will surprise you.

In baseball, there's been a movement called sabermetrics which aims to uncover truths about the game through statistical research. Using the copious numbers provided by the game, analysts have come closer to discovering the true value of different baseball events, the right time for certain strategies, and better ways to assess player performance.

It's time to take that attitude to beer.

We'll still be about enjoying beer here at BeerGraphs. "Get your nose out of the spreadsheets" is a common refrain from old-school baseball fans to the sabermagicians, and I suppose some old-school sudsers will feel the same way about us -- but the point is not to sterilize beer, but to celebrate it. And really get into the relationship we have with our favorite adult soda.

If beer is culture, then BeerGraphs will narrow that down and talk about Beer Culture specifically. With two blogs and some leaderboards, we hope to have different ways to help readers enjoy beer more readily. BeerGraphs the blog will have in-depth pieces on the science of brewing, interviews with important people in brewing, and then analysis from our crack team as they try to uncover biases, trends, and quirks in our beer ratings. BarelyBeer the blog will feature, ostensibly, beer reviews. But each of our columnists will have control over their personal brand -- they'll be writing fun pieces that happen to include beer reviews instead of the other way around.

Our leaderboards will be a work in progress. The more data we acquire, the more we can pass on to you. We have more data coming shortly -- beer aspects like IBU and ABV -- and we have long-term hopes for even more. Today, we debut with a peek at our leaderboards to come: The BeerGraphs Top 100.

We also want to be a hub for crowdsourcing problems: trading, finding good bars and stores for craft beer, entering in reader reviews. We'd like to give you fodder for barroom discussions about the true ranking of pale ales. We want you to be able to have those barroom discussions here if you like. Perhaps we could find a user-friendly spin on the beer festival. We will have meetups. This will be nerdy. This will be exciting.

Likely, if baseball is any guide, the leaderboards will be the most popular early feature. And our new beer rating -- Beers Above Replacement, or BAR -- might be fodder for debate.

We will refine the metric over time. We invite your input, whether it's on Twitter or in the comments. We hope to make this better. My next post will posit an early question about a fundamental concept in BAR, for example, and hopefully after a short series on the subject, we'll be able to add a new wrinkle to the rating early on it's life.

The BAR rating will be tackled in the upcoming week by the same analysts that helped come up with the initial metric. So I'll let them explain how we found a way to reward beers for being drunk a lot, and yet also to reward beers for being good. This is an important part of beer drinking, the relationship between availability and quality, and we'll be working on that for a while.

But I want to talk about an aspect of BAR that was basically, for me at least, the genesis of this entire idea.

Our ratings will be style-specific.

A five-win IPA might not have the same raw ratings as a five-win Pilsner. Here's why we think that's a good idea.

Take a whirl around the top of the leader boards for the different styles at Beer Advocate, and you might notice something:

Is it fair to Pilsners that they don't have a 100? How about Bocks? Their best doesn't even register above 90. Let's say I like Bocks, and Pilsners, am I to take the Pilsner because when I log on to it and check it against the Bock, the average Pils wins easily? And why should we denigrate the Pilsner drinker? I love Pilsners! The craft beer world loves dark beers, but let's not poo-poo the rest of the platter.

There's another way of looking at the same numbers. Here are the top 15 styles of beer in Untappd's database.

Num Style Average
1 Imperial / Double IPA 3.654723
2 Imperial / Double Red Ale 3.636921
3 American Imperial / Double Stout 3.54715
4 Imperial Oatmeal Stout 3.494345
5 Russian Imperial Stout 3.478876
6 Belgian Quad 3.448507
7 Imperial / Double Black IPA 3.448224
8 Root Beer 3.444068
9 Flanders Red Ale 3.437134
10 Gueuze 3.435954
11 Imperial / Double Porter 3.418754
12 Milk / Sweet Stout 3.356179
13 Cyser 3.34609
14 Irish Dry Stout 3.344916
15 American Wild Ale 3.338067

Three of the top five beer styles are stouts. The top seven beers are all heavy in one way or another. Yes, the first is a pale ale, but it's not a Pale Ale or even an Imperial Pale Ale, it's the heaviest, bitterest, strongest version of the style that rises to the top. Belgians come in sixth, but it's not just any Belgian, its the quad monster. You real have to get down to numbers ten and eleven to get something 'lighter,' but those two beers are both sours. And then a porter. And then a stout. The fifteenth-rated beer, American Wild Ale, is the lightest on the list, as an American Saison basically.

I love Saisons. Of the 316 unique beers I've had since I first logged on to Untappd last January, 48 have been American IPAs, 32 have been Imperial IPAs, and 33 have been Saisons. I also like Saisons more than the average suds soiffer: I've even given them an average rating of 3.97, while other untappd users have given those particular beers a rating of 3.76 collectively. I even gave five stars to the Saison Imperiale from De Proefbrouwerij while other raters gave it a 3.67. So I maybe I'm not unbiased when it comes to the style.

And yet on Saisons carry the 6oth-highest average rating (3.07) on untappd. On Beer Advocate, only one Saison merits a 100 rating, while *four* American Imperial Stouts have 100s. I've had 11 of those top Stouts, so I don't mind them, but in my world, there are as many perfect Saisons as there are Double Stouts. Should I just ignore the lower ratings? Should I follow the herds and drink more double stouts? What about my feelings about popular styles?

No, I think it makes sense to rate a beer within its style, and that's why I like BAR already, even in its initial, flawed state. You might have as many seven-beer-win Saisons as Double Stouts, because we tried to make it so. And if the beer winds shift from Double Stouts to Saisons some time in the future (not such a crazy idea), we'll be ahead of the (win) curve at BeerGraphs.

If you're in the bay area, come to our meetup this evening!

Interested in BeerGraphs? You can contact us here.