You love the blog, so subscribe to the Beervana Podcast on iTunes or Soundcloud today!

Monday, October 06, 2014

GABF Analysis: Five States Won Half the Medals

Last update, next day.  Two very interesting analyses came out today that are FAR more interesting than what you'll find here.  Both incorporate the numbers of beers entered per state, which make for a much richer and more accurate picture.  Stats geeks (and Oregon homers) will love them--they're sort of like the sabermetrics of beer.
___________


Update, 11:55 am: Geoff Kaiser posted an incredibly interesting--and potentially revealing--comment about the number of breweries pouring at the GABF.  His figures are probably a lot closer to representing the number of entries per state.  It could explain a lot.

___________

As you are probably aware, the Great American Beer Festival concluded this weekend, keyed by the announcement of the annual awards and a storm of joyful tweets.  As usual, I tuned in to see how the locals did, delighted to see some impressive and unexpected results.  Tiny Arch Rock Brewery in Gold Beach took home a gold in the kellerbier category.  (You first read about that brewery and that beer--" the lager, predictably my favorite, is a rustic, slightly hazy beer more in the Czech tradition than German"--right here on Beervana.)  Barley Brown's once again demonstrated dominance in the hoppy categories (CDA, international pale ale, and gold in fresh-hop), and Portland's own Breakside took honors with a gold in the most coveted of awards, American-Style IPA (a fest-record 279 entries).

My own homerism got me thinking: how are medals distributed?  This was the first year in which a new registration system that vastly expanded access to the Fest.  In 2013, many breweries applied and were denied entry--and about 700 managed to get beer into the fest.  This year, the Brewers Association made room for up to 1400, and over 1300 submitted beers.  That's actually a huge number--right about half of all breweries in existence at the beginning of the year.  In all, an astounding 5,500 beers were reviewed.

I often whinge about the ever-growing number of categories, which ensure an ever-growing number of medals.  There were something like 273 medals possible, but judges didn't award medals in some categories, and in at least one category--Koyt beer--there were no entries.  (Is this proof that you can have too many categories?  I mean, koyt beer?)  Overall, 268 were awarded, which seems like a lot until you consider that only one medal was awarded for every 20 entries.  (Harvard's acceptance rate was higher.)  Of the 1309 breweries in the competition, just 234 managed to medal (17.9%).

And yet for all that, when you look at the medals awarded by state, you see anything but an even distribution.  We don't have numbers on how many breweries entered per state, so the figures only tell part of the story--but it's quite a story.  Just five states won half the medals, and just two states--Colorado and California--account for almost a third.  This is especially impressive for Colorado, which has less than half as many breweries (180ish) as California (381).  Here are the top ten, which constitute nearly two-thirds of all the medals won; the other 40 states managed to collect just 35%.


To add a bit more detail, here's a list, with the approximate number of total state breweries included (and keep in mind, I have no numbers for how many of the state's breweries actually entered the competition).

Medals Won
CA - 46 (approximately 318 breweries)
CO - 39 (180)
85 cumulative medals (32% of total)

OR - 22 (181)
TX - 16 (96)
PA - 12 (108)
135 cumulative medals (50% of total)

IL - 9 (83)
WA - 9 (200)
NM - 8 (31)
VA - 7 (61)
WI - 7 (90)
175 cumulative medals (65% of total)

Other states receiving at least five medals: MI (6), MN (6), MT (6), NC (6), MD (5), OH (5), UT (5).

I would be leery about over-interpreting these results.  Nevertheless, it's impossible to avoid the conclusion that a very few states seem to be way out in front in terms of making making medal-earning beers.

8 comments:

  1. I have decided that medal numbers must increase with brewery openings to ensure a statistically sound likelihood of everyone winning something regularly to ensure the PR promise of we are all great is proven.

    ReplyDelete
  2. "We don't have numbers on how many breweries entered per state..."

    Yep, until the BA decides to start releasing that info, we can all make as many assumptions as we want...but we're just guessing.

    The best we have to go by is the # of breweries pouring at the fest, and here is that breakdown for the top states...:
    State Total
    CO 134
    CA 123
    TX 44
    OR 32
    IL 31
    WA 25
    VA 22
    PA 21
    OH 18

    ReplyDelete
  3. Geoff, that's incredibly interesting. Where'd you find those numbers?

    ReplyDelete
  4. It would be more interesting to normalize these to median salaries, number of distinct metro areas, total population, and other variables to see if we can mine any distinct correlation.

    The state that interests me above is WA. (and to a minor extent Colorado, but there is a host effect in play).

    WA however begin so different from Oregon is just surprising to me.

    ReplyDelete
  5. What was really interesting to me is that some of the breweries that medaled from Colorado were a year or two old.

    ReplyDelete
  6. Jeff - on the GABF website, there is a drop down table of all the breweries pouring on the floor. You can copy the fields into Excel and do a pivot table from there.

    I "think" I have an even better set of data coming from the BA that will give me the actual number of participating breweries and number of beers entered for each state. No details on the specific breweries/beers though; the BA likes to keep that secret, which I get.

    I'll let you know what I get.

    ReplyDelete
  7. As implied, until we know the actual number of breweries entered from each state, the numbers are pretty meaningless. They're probably pretty impressive, but that can't be said with any certainty at this point.

    ReplyDelete