Medals and math IV — two in a row seems like it means something…

CORRECTION — the original version of this post incorrectly said that ‘Chance, Luck & Magic’ was the overall Champion Beer for the second year running. It wasn’t; Burkes Brewing ‘Unforgiven’ Porter won in 2021. Thanks to Michael Donaldson for noticing and letting me know. The data here is unaffected, but some of my commentary was thereby off.

An opened bottle of Garage Project's 'Chance, Luck & Magic' beer
A cork popped by way of congratulations

After this year’s Brewers’ Guild of New Zealand Awards — the BGONZAs, to their friends — the headline result was unusually clear: Garage Project1 just absolutely smashed it. Their ‘Chance, Luck & Magic’ took out the Best In Show award, prompting me to buy a $49 bottle of beer (which was suitably delicious, I should note), and they won the Champion Large Brewery title (for the second year running, no less) in spectacularly unambiguous fashion, since at least two other breweries would need to merge and pool their winnings to come close to G.P.’s medal haul.2 After that, it might seem superfluous to dig in the weeds of the data for other stories lurking in the details, but I have my traditions, and I am undeterred — and I think there’s still some things worthy of a little more attention.

So, as I do with every fresh batch of results now, I’ve calculated two ‘scores’ for each brewery that are extra ways of looking at how well they performed. MPC is medal percentage: how many beers entered won a medal (any medal). PPE is points per entry: if we say a gold is worth 3 points, a silver is 2, and a bronze is 1, then we can work out a sort of batting average. Bigger numbers are better for both, and the competition averages this year were MPC 61.1, and PPE 1.02. If you beat the odds on either of those (or better yet, both at once), you are, in a very real sense, doing well, even if you don’t earn a title or make the headlines.3 For the breweries that entered ten or more beers (which is about half of them, and cuts out the noisier end of the data while still including all the champions), the table4 looks like this:

Performance at the 2022 BGONZAs, by breweries with ten or more beers entered

n = number of entries, MPC = medal percentage, PPE = points per entry, G/S/B = individual medals

Garage Project (Champion Large)3387.91.709911
Behemoth (Champion Medium)4566.71.1351114
Three Boys (Champion Exhibitor)1392.31.92372
8 Wired1770.61.35354
Three Sisters (Champion Small)1936.80.74313
Good George1384.61.62263
Shining Peak1566.71.20244
Kainui (Champion Micro)1464.31.07225
Beer Baroness131001.69175
Fortune Favours1346.20.77123
Deep Creek20901.35099
Double Vision16500.69035
Ground Up1233.30.58031
Fork & Brewer1833.30.44024
Hop Federation10600.80024
Sprig + Fern20500.55019

I commented a little on the new(ish) ‘Champion Beer’ award when it was first awarded, and I honestly still don’t really get it. They take the gold medal winners from a category (like Stout & Porter, or International Lager) and taste them against each other to award a ‘Best in Class’ trophy,5 and then all of those compete for the overall title. The first step there makes sense to me, but I struggle with how to judge this pilsner-as-a-pilsner against that porter-as-a-porter. It feels like it must eventually come down to voting on individual preference, which isn’t what this whole thing is supposed to be about. While it’s early days for pattern-spotting among the winners, here, I can’t help but notice that three out of four of them so far are vaguely-Belgian-inspired,6 which might say something about the backgrounds and biases of the judges, and what counts as prestigious around the final table.

That said, G.P.’s ‘Chance, Luck and Magic’ has the extra honour of being a trophy-winner two years in a row, which certainly shores up the idea it was something special. Winning two gold medals in successive years is rare enough,7 happening only three times at this year’s competition.8 By necessity, taking two trophies in a row is even more difficult, and I’m definitely keeping an eye out for that kind of serial success. The Australian International Beer Awards have a category (Consistency of Excellence) that specifically calls out and celebrates beers that win gold three years in a row. I think that’s an idea worth copying.

Awards storage (Garage Project, 10 October 2017)
Garage Project have outgrown the old awards bin, in more ways than one.

At the brewery level, though, Garage Project aren’t the first to win successive championships. Even just in recent years, Lion and Fork were both back-to-back champions (Large and Small, respectively) in 2018 and 2019, and Three Sisters are this year’s Champion Small Brewery after being 2021’s Champion Micro Brewery, which is a very neat achievement indeed.9 But G.P. didn’t just spam the awards with beers to collect their unprecedented nine gold medals, their 33 entries (the third highest total in the running) earned them extremely high MPC and PPE scores, well into the top end of the field for each, and both are significant improvements on their already-impressive 2021 results. So while a repeat championship has been won before, this repeat still stands out pretty clearly. Among the various Champion Breweries of the last five years, their performance is only topped by Liberty’s remarkable turn in the Medium tier in 2019 (with MPC 100 and PPE 2.18).

Other than Three Boys,10 this year’s Champion Exhibitor (which is essentially awarded on the basis of PPE, with a few caveats), the Medium, Small, and Micro Champions this year are unusually off the pace in terms of MPC and PPE; if you sort the above table by either metric, all three drop out of the top ten, which hasn’t happened in previous years. It looks like Three Sisters enter many more beers than other breweries of their size, and given the way the championships are awarded (essentially a count of golds, with ties broken by lesser medals) it’s paid off for them two years running. Their MPC and PPE aren’t nearly as bad as Sam Adams / Boston Beer Company regularly had while winning the Champion International Brewery year after year,11 but still. Behemoth, 2022’s Champion Medium Brewery, improved significantly on their last year’s scores but also seemed to take the gong by playing the numbers game and again entering the most beers of any competitor. This is the competition as designed, after all, I just highlight this kind of thing because other systems are possible and a metric like mine can reveal some pretty impressive performances (I’m looking at Beer Baroness, Good George, Deep Creek and Panhead in particular) that otherwise go unheralded. In Australia, this year’s Indie Beer Awards are switching to a new method for crowning their size- and state-based champions, with a focus on entrants’ “overall performance and not just their top beers”. I’ll be watching that with interest — and am very curious to peer under the hood and see their actual formula.

The other thing that leapt out at me, this year, was the breakdown of performance at the awards by style category. The Guild included a handy table of totals in the results booklet,12 but when you put it into medal percentages and batting averages, there’s some pretty stark differences in how much of a fight there was over the various trophies. The new ‘No & Reduced Alcohol’ class (an excellent addition, I think, and certainly better than trying to shoehorn them all into ‘Experimental’ or making them compete against full-strength versions of whatever style) is unique in having zero gold medalists despite a healthy number of entrants. Meanwhile the other new category, ‘Juicy / Hazy’, was easily the most crowded race with nearly twice as many beers entered here as there was in the non-cloudy IPA bracket — and despite its novelty, results within it were pretty much on par with the rest of the competition. I’ve included the whole table below for the supercurious, and you can easily see that the variances are pretty wild both in terms of how many breweries were vying for different trophies and the average performance in those subsections.

And… that’s all I’ve got for now. There’s a lot of data, especially after five years of publishing the results like this, but there’s always the risk of starting to hallucinate patterns where there are none if you gaze too long into the spreadsheets. Hopefully I’ve surfaced a few useful trends or bits of trivia, if nothing else — additional observations are, as always, welcome.

Style category contestation at the 2022 BGONZAs

brewers contesting the category, total beers entered by all breweries, MPC = medal percentage, PPE = points per entry, G = total golds, GPC = percentage of beers winning gold

Amber / Dark Ale181973.71.0015.3
Amber / Dark Lager161877.81.1715.6
British Ale101776.51.41211.8
European Ale233770.31.24616.2
Fruit & Flavoured399970.71.1599.1
International Lager395259.60.9223.8
International Pale Ale355457.40.8323.7
Juicy / Hazy5613557.80.99107.4
No & Reduced Alcohol202962.10.8600
NZ Styles4710356.30.9076.8
Specialty & Experimental428279.31.341113.4
Stout & Porter395563.61.1159.1
Wheat & Other Grain151643.80.94318.8
Wood & Barrel Aged1524751.4214.2
Grand Total7881163.41.05647.9

  1. I keep saying this, despite never knowing if anyone cares or if it’s really relevant, but: I worked there for several years (several years ago) and count a decent chunk of their staff as friends, for the record.
  2. The competition is split into four tiers based on total brewing output, but who is in which isn’t made public and we’ve got limited data to draw from. That said it sure looks like Garage Project outperformed the next three  or four (maybe more) of their ‘Large’ competitors combined, this year.
  3. And — to be clear — if you win a trophy or championship, whatever your MPC or PPE, you are also doing well and playing the game as it is given to you. I just like looking at things in a few different ways, not least because it makes it easier to imagine how the game might be refined.
  4. Thanks, again and always, to the ridiculously convenient Tablepress plugin.
  5. The trophies aren’t part of my MPC / PPE analysis at all. They do kind of operate as a supergold — though it’s possible for a silver medal beer to win one, if there are no golds in a category, as happens sometimes — so maybe one could count for 4 points, or 5? And some classes are more hotly contested than others, should that factor in? How? For once, I decided not to overcomplicate things.
  6. The exception being the Burkes porter I admitted to forgetting about, above. The others were Deep Creek’s Diep Kriek (2019), and Moa’s ‘Southern Alps’ (2018), a White IPA which was at least recently still branded with ‘Judged NZ’s Best Beer’, despite failing to earn a medal of any kind since, which is a good lesson in caveat emptor when it comes to awards (and probably a few other things).
  7. You might assume (I know I did) that given the nature of the beer business, it’d be rare to see the same beers entered year after year, with the churn of new releases that dominate our attention. But I found 336 different beers that were in the running in at least two of the last three competitions, and 81 of those were in all three.  Their producers (and fortunes) varied wildly but that’s perhaps a story for another time.
  8. One of the others was another Garage Project beer. But the third was Rheineck, which is worth pointing out if people mistake these for “craft beer” awards, or get snobby about big breweries not making “good beer”.
  9. Michael Donaldson — himself a trophy-winner again this year — has a nice write-up of their progress.
  10. Yes, of 2022’s four Champions, one is called “Three Sisters” and another is called “Three Boys”. Brewery names have definitely reached that stage that pop punk bands also quickly hit around the turn of the millennium where they tend to blur together and clump around a few tropes. Wellingon also has Whistling Sisters, Melbourne has 2 Brothers. My home town has Panhead and Boneface, and a few hundred metres from where I live now, there’s Bonehead; I’m almost tempted to start Panface, just to complete the set.
  11. The stark weirdness of which was my main takeaway the very first time I ran these numbers.
  12. There were a few typos in the original booklet that mucked with my original data entry run, but the Guild helpfully provided me with an export from their internal working-out spreadsheet which sped things up considerably. Granted, it’s still about a month after the awards were announced, but I’m beating my previous delay in getting something up here by quite a margin, right?

One thought on “Medals and math IV — two in a row seems like it means something…”

Have at it: