GABF Medal Results vs. Number of Beers Entered – A Closer Look

October 7th, 2014 · 11 Comments · General Beer News

Warning: this post is for beer or stat geeks only. It is also probably a waste of time, but as someone who deals with analyzing numbers for his day job I couldn’t help but put a little effort into this. 

2014 GABF Medal Ratio2A post by Jeff Alworth over at Beervana yesterday got me thinking. Jeff took a look at the number of medals won by each state at this year’s GABF and then wondered about the distribution of medals and if the results mean that “…a very few states seem to be way out in front in terms of making making medal-earning beers.”

Looking at the medal results, the top 5 states, CA, CO, OR, TX and PA, took home exactly 50% of the medals from the 2014 GABF competition. Without taking anything else into consideration, that is certainly impressive. But, how many beers did each state enter into the competition? That is the key piece of data that might give us some insight into the numbers.

I reached out to the Brewers Association for help, and while they understandably don’t want to divulge exact details on all of the beers entered into the competition, they were happy to provide the total number of beers entered by each state. That gives us a meaningful comparison of the number of beers entered vs. the number of medals won.

The table above (click on it to see a larger image) shows a comparison of the ratio of the number of medals won vs. the number of medals entered, as well as a comparison of the total % of entries vs. the total % of medals for each state.

So, what sticks out?

  • California submitted the most entries, but they didn’t win any more medals than they “should have”. They accounted for 17.2% of the entries and 17.2% of the medals, which is a perfectly average win %.
  • Oregon, Colorado and Pennsylvania had a substantially higher (>= 1.0%) “win percentage” than most states. Oregon’s showing was particularly impressive, winning 8.2% of the medals but submitting just 5.1% of the entries. To me, that confirms what Jeff A. was getting at: Oregon breweries do indeed kick more ass than average.
  • Arizona, North Carolina, Florida, New York and Ohio had a substantially lower (<= -1.0%) win percentage than most states. Arizona submitted 2.1% of the total entries but took home just 0.4% of the medals.
  • I’ll fully admit that the main reason for me doing this is wondering why Washington “only” took home 10 medals when we have the 2nd most breweries in the US. It turns out that our breweries submitted 214 beers total, good for the 6th most of any state. The 10 medals won represents 3.7% of the total medals vs. submitting 3.8% of the total entries. So, the win percentage is just about average. A discussion on why WA breweries don’t enter more beers is probably worthy of a whole separate thread.
  • Looking at just gold medals, WA actually led the way in “Gold Medal Ratio” for the 18 states submitting at least 100 beers. WA won 6 gold medals for a 2.8% ratio vs. # of beers entered. PA was next best at 2.5% and then OR at 2.4%. CA had the most gold medals with 15, which equals a 1.6% gold medal ratio.

Let me stress that this is all to be taken with one big ol’ grain of salt. This is what numbers/stats geeks do….lay out the facts that the numbers present, but let people come to their own conclusions. If a state had a low win %, it doesn’t mean their breweries suck. Maybe they entered more beers in the super competitive IPA category, etc… Maybe states with a higher win % cherry-picked the categories with less entries. Or, maybe this is a competition judged over a grueling week by 200 volunteers that can’t see straight by the end of it. Also, the number of entries and winners by state are certainly only “directionally correct”. Some beers reported as from CA might actually be brewed somewhere else, and so on.

 

 

Tags:

11 Comments so far ↓

  • Drew

    Very nicely done Geoff, I being a numbers geek even more then a beer geek really enjoyed this post. Drew

  • Jeff Alworth

    Amazing! Fantastic! Way better than my meatball numbers! I think you’re understating the reliability of the numbers here. They’re massive samples–even state by state–so there’s no issue of significance. The only (small) caveat I’d add is that it’s a snapshot from just one year. If a homer wants to make the case their beer is better generally–as opposed to just really good that year–you’d have to see the trick repeated over time.

    Great job.

  • Bart Watson

    Glad to see other people digging into this! Here’s my analysis that controls for the number of entries in each category:

    http://www.brewersassociation.org/insights/gabf-medal-winners-analyzed/

    WA is still pretty much in line with what would be predicted solely based on entries and the categories WA breweries entered.

  • Brad H

    Do you have the numbers for the rest of the states? Can you post those?

  • Kaiser

    Hey Brad – I’ll send you the full list to the email you registered with (which looks valid).

  • Kaiser

    Jeff – I have actually already asked for data for the past 10 years of competitions, but I don’t think that is going to happen…we will see.

    Bart – that’s great stuff. Thanks for putting that together. Do you happen to have any similar comparisons from past years?

  • Jeff Alworth

    Geoff–if you’re able to pull that off, you’ll deserve some special recognition for bad-ass blogging.

    Bart–thanks for putting your numbers together. They’re really fascinating and equally revealing. (Medaling in American IPA, as Oregon has done for the last three years, really helps goose our stats!)

  • Russell Everett

    Great analysis Geoff!

  • Bart Watson

    Haven’t done previous years – though it’s something I’ll plan to look at going forward.

  • Mike Lava

    I, too, would love to get the data on the rest of the states. I mapped the winners and would love to include your data as part of the visualization.

  • Mark Hood

    Very interesting and a valuable way to look at it Geoff, thanks. I agree with Jeff, it’s a big enough sample size to be meaningful

Leave a Comment