Friday, November 21, 2014

The ... best ... beer ... in ... the ... nation! (and in your state)

What are the best 'craft' beers in every state of the nation (and the District of Columbia)? Who are the best breweries in each state? What are America's top 'craft' beer styles? Which of the states are best for 'craft' beer?

To find out, Bryan D. Roth, at his blog, This is Why I'm Drunk, crunched October 2014 data from Beer Advocate —a crowd-sourced online beer-rating site— using the popular site's algorithm of weighted measurement for beer ranking.

He found answers —some expected, some surprising— and quantified some current trends.


To establish a minimum quota of scores that any beer, in any given state, needed to be considered as a top beer (somewhat analogous, say, to Major League Baseball's requirement of 506 plate appearances for any player over one season to be eligible for the batting title), Roth averaged the number of ratings for every beer in that state. From there, he referenced each qualifying beer's weighted rank and came up with each state's top ten most highly rated beers.

To decide upon the top brewery in each state, Roth looked at the number of times a brewery's beers appeared in the top ten, and, in the case of a tie, chose the brewery with the higher total ratings in the top ten (or, if needed, tossed a coin).

For instance, in California, the top brewery was Russian River; in Colorado, Great Divide; in Oregon, Deschutes; in Texas, Jester King; in Pennsylvania, Tröegs; in Vermont, Hill Farmstead.

Here, in the 'DMV,' the tri-'state' home of YFGF:
  • In Washington, D.C., DC Brau took top honors with seven of the ten top beers, and the number one beer, On the Wings of Armageddon, a double IPA.
  • Flying Dog placed first in Maryland, with four of the top ten beers. The most highly rated beer, however, belonged to Stillwater Artisanal, for its Gose Gone Wild.
  • In Virginia, the results were less cut-and-dried. Hardywood and AleWerks dominated the Commonwealth's ratings, producing four beers apiece in the top ten. Hardywood's most highly rated beer, Gingerbread Stout, placed second, while AleWerks' milk-stout, Café Royale, placed third. AleWerks, however, nudged out Hardywood as tops in the state by virtue of a higher total of weighted ratings. The number one beer was Resolute, a bourbon-barrel 'Russian Imperial' Stout, brewed by Three Brothers.


Roth ranked the top beer from each state (and Washington, D.C.) by giving each state equal status despite population (as do, by the way, our U.S. Senate and presidential Electoral College), and then using the BA's weighted rank for the top beer in each state. The top 3 beers, nationally, by state, were:
  • Hunahpu (barrel-aged Imperial Stout), from Cigar City, in Florida. 11% alcohol-by-volume (abv).
  • Kentucky Brunch (Imperial Stout), from Toppling Goliath, in Iowa. 13% abv.
  • Heady Topper (Double IPA) from The Alchemist, in Vermont. 8% abv.
At the bottom, at number 51, was Pile of Dirt (porter) from South Dakota's Crow Peak Brewing (6% abv). But remember, using these metrics, that beer is still 'better' than 90% of the top ten beers from all the other forty-nine states, plus D.C. See the full list: here.

[Here, in the DMV, Stillwater's Gose Gone Wild finished at position number 21; Three Brothers' Resolute at 23rd; DC Brau's On the Wings of Armageddon at 27th.]


Again giving equal importance to the best ten beers in each state (disregarding sales or population), Roth pie-charted the nation's top beer styles.

IPA (India Pale Ale) —at 18% of the total— is, of course, the most popular beer style; combining its percentage with that of DIPA (double IPA), the figure becomes 31% of the total. Pumpkin? At 1.5%, it's near the bottom of the heap. Well, that's not exactly fair on my part. Pumpkin beer is, after all, only seasonal —even if that season seems to run from Independence Day through Thanksgiving. There's also a catch-all "Other" category, in which there are other, less popular beers. 'Popular' might not be the correct word. Roth uses the phrase "lauded and sought after."


What about the correlation between alcohol and scoring preference?
The top-25 clearly has a higher ABV against both the bottom-26 and the overall average, not to mention a rather large difference in Beer Advocate’s weighted ranking score. I was rather surprised the the bottom-26, even though it’s the “best of” what states had to offer, barely beat the “all beer” average in weighted rating.

Roth looked at the average alcohol rank of the top ten beers of each state with the average weighted rank of its top ten beers versus all other states. They seemed to correlate, but not one-to-one. The average alcohol percentage of all 506 beers Roth surveyed was 8.1%; the average of the top beers from each state was 9.65%.

California was tops in both average of alcohol percentage of its top 10 beers (11.18%) and of the average weighted rank of those beers. Delaware, 2nd in alcohol (11.174%), was 27th in weighted rank, and Alaska, while 3rd for alcohol (11.08%), was ranked 15th for the average of its top ten beers. Only those three —California, Delaware, and Alaska— had alcohol averages surpassing eleven percent. Interestingly, Vermont, while number two, nationally, in average weighted rank of its top ten beers, was only 26th in average alcohol level (8.1%).

A more pronounced alignment of alcohol and score did occur at the bottom rung of the state rankings. There, alcohol levels were markedly lower, as were weighted ranks. The lowest alcohol average belonged to South Dakota (5.21%) whose weighted rank position was 50th out of all states. Obviously, strength matters.

[Washington, D.C. was 41st among all states in average weighted rank and 43rd in average alcohol percentage: 6.17%. Maryland was 28th in average weighted rank and 22nd in average alcohol percentage: 8.035%. Virginia was 23rd in weighted rank and 12th in average alcohol percentage: 9.39%.]


What if a savvy brewer were to look at all these data, and want to brew a beer to conform to the prejudices of Beer Advocate's voters? Well, then, Roth has a recipe for her:
  • It has to be 18 SRM [a measure of beer color], the average number for all 506 beers on the “best beers” list.
  • It has to have an ABV of 8 or 8.1 percent, per the average of the list.
  • It must be hopped like an IPA.
  • From there, two optional aspects:
    • It can be hopped like a DIPA.
    • It can be barrel aged.

Get to it, brewers. Brew the ... best ... beer ... in ... the ... nation!

Roth's analyses —fascinating stuff— bring some data-driven sense back to the 'craft' beer zeitgeist. This series —a lot of work— demonstrates why Roth was deserving when the North American Guild of Beer Writers selected his blog as one the three best in the nation for 2014.


No comments:

Post a Comment

Comment here ...