Quantcast

What Makes A Beer Active

Michael Donato, August 13, 2015 -   

The beer world is awash with one-off brews, collaborations, experiments, retired beers, and old vintages. Most people aren’t interested in what the best beer is all time, or what’s the best saison ever brewed, they’re interested in beer that they can get on the shelf right now.  

It doesn’t add much when beers that haven’t existed in two years show up in top beer lists. Even our leaderboards have the 2012 vintage of Bourbon County Brand Stout at number 10, something that’s not particularly helpful to know. Sure, a lot of people have these cellared, but it’s not exactly a beer that you can run out and get. It’s not active. Wouldn’t it be nice if we could filter out beers that aren’t active?

Turns out we can, but it’s not that simple. The relative volume of beers a brewery puts out makes this an uneven exercise. I initially tried looking at checkins over the last three months as compared to the whole, but for the large beers that have been distributed regularly over the roughly two years we have data for, that represents merely 12.5% of total checkins and minor fluctuations in recent data make it look like it’s at least less active. Clearly not true for beers like Lagunitas IPA. 

Next I tried to go back six months. Let’s take a look at the number of check-ins in the last six months as compared to the total number of check-ins for that beer. Then let’s find a percentage that eliminates the inactive beers and provides us with a nice and current sample of beer to work with. This presented problems as well: There were still many seasonal type beers sneaking into every percentage I tried. Beers from six months ago that were brewed once aren’t generally available, or fresh, today.

I’m only looking at beers with 100 check-ins as my floor.  It’s pretty hard to judge if a beer is active or not when it only gets a few check-ins a month. First I looked around the 5% mark--so beers with 5% of its checkins within the last six months. Here’s a subset:

Brewery

Beer

Check-ins

Recent %

Collaboration: Sierra Nevada / Ballast Point

Electric Ray

3498

5.002858776

Shipyard Brewing Company

Bourbon Barrel Aged Smashed Pumpkin

380

5

Amnesia Brewing

Desolation IPA

236

5.084745763

Boulder Beer Company

Honey Saison

1269

5.12214342

Brewery Ommegang

Scythe & Sickle

4887

5.054225496

This isn’t particularly helpful. Scythe & Sickle and the Pumpkin are clearly seasonals, both harvest ones that are almost active again. The collaboration was obviously a one off that is clearly defunct. The other two represent a conundrum. Boulder’s Honey Saison is indeed seasonal, usually released in February. That means almost all of it’s check-ins should be active ones, but clearly not showing up that way. Amnesia’s IPA is a year round beer, but perhaps it’s limited footprint is making it approach inactive status. Should these beers be culled? or saved? We could drop back to 4%, but that shows many of the same problems. 

One hurdle to work with here is seasonals. You wouldn’t expect many Oktoberfest or Pumpkin beers to show up in June. We might still want these to appear as active since they’ll be available again soon. They’re not really inactive, they’re just hibernating.

Perhaps we can tweak the end points of our query to try to capture all seasons, but between seasonal creep and limited release beers not adhering to the calendar specifically, it’d be hard to accurately capture that.  Perhaps seasonals occasionally being inactive and then suddenly appearing again is exactly how it should work. Active beers, right now, is a different question than which beers are active in 2015 as a whole. 

We need to accept a bunch of false positives, whether it’s a flurry of popularity for a somewhat older beer, or fading seasonals clinging to life.

So I changed my query to stop looking at the whole database and only looked at the last thirteen months and compared it to the last 4 months. I went back to the start of July 2014 for the total sample, and the start of April for the active period. That’s enough to factor in every seasonal, but combined with our hard floor of 100 check-ins, helps eliminate more beers not brewed anymore.

More than 5% of Three Floyd’s Dark Lord (2013) have been consumed in the last three months and there are a bunch of 2013 Cantillion at a higher percentage than that. Guinness, the beer with the most check-ins over the past year, is less than 6%. So let’s keep the percentage low. Let’s only eliminate the beers that have 1.5% or less of their check-ins over the last four months. This will include a ton of late in the year seasonals, but they’ll be back soon enough. 

This drops beers like Sierra Nevada Alt Route, Founders Sweet Repute, and Abita Pecan Harvest. Redhook Pilsner, a retired beer, gets bumped. So does Lagunitas A Little Sumpin' Wild Ale, which is an August-October seasonal. Despite 60 check-ins, we lose Sierra Nevada’s Bigfoot (2013) which suggests a lot of people tend to drink these big beers a year later, but not so much 15-18 months later. This will be a good one to keep tabs on with an eye towards understanding how vintages are consumed.

Surprisingly, Bigfoot (2012) which is up near the 1.6% mark, makes the cut. As does Stone’s 6.6.6. Vertical Epic, though 11.11.11 gets the axe. Christmas Ales, pumpkins, smoked porters and various winter and fall themed beers fall by the wayside, but if we venture higher than 1.5% we lose a bunch of smaller, but available, beers like Peace Tree’s Hop Wrangler or Starr Hill’s Starr Pils. Also just above 1.5% are beers with dwindling interest but not availability like Budweiser Black Crown or Coor’s Third Shift Amber Lager. Stone’s recently retired Levitation Ale is on the cusp, but remains active. 

This process is probably not completely without its false negatives, though I checked a couple of dozen beers and only found seasonals and limited or rotating release beers that were currently not being brewed.

Brooklyn Brewery’s Brooklyn Pennant ‘55 Ale comes in with 41 check-ins compared to 4312 total. This is a year-long beer that’s been around for years and I can find no indication that it’s been retired, or down-sized like New Holland’s Rye Hatter which was only brewed for draft this year making it seem retired by comparison. Something appears up with Brooklyn’s Pennant Ale, as there is vague allusion to it being retired on Twitter, but nothing official. Perhaps my research is breaking news here. 

So I asked.

Breaking beer news. And on the 60th anniversary too!

I’d like to add one more filter here. The larger a beer’s volume, the larger the amount of straggling bottles. Take Blue Moon Gingerbread Ale as an example. This Coors product has 3158 check-ins over the last 13 months, but only 43 in the last 4 months. Despite being above the percentage threshold, this beer is clearly not active. Let’s remove any beers with at least 1000 check-ins over the last 13 months that still have a recent volume of less than 50. We don’t want huge distribution seasonal beers sticking around over their peers based on stragglers from variety packs. This would also nix beer’s like Yuengling Bock and Shmaltz’s He’Brew Jewbelation Reborn.

There are just shy of 600,000 unique beers in our database. This process eliminates a lot of the noise involved with looking at the entire database, and allows us to look at mainly beers that are actually on the shelves, or will be soon. Filtering for just the last 13 months and 100 check-ins brings us to 30,200, and eliminating the inactives of that list cuts off another 6,200 or so.

That brings us to a rather manageable 24,000 beers that can be deemed ‘Active’. This number will fluctuate some as bigger seasons add new beers while smaller ones drop off temporarily, and of course the ever expanding number of breweries will add more and more active beers. For now, 24k is what we’ve got. 

Michael can be found on Twitter and Untappd and took that header image of faded beer bottles a few years ago. The place has since burnt down. You can also email him at beer@ceetar.com.

comments powered by Disqus