13 Nov 2009 @ 11:54 AM 
 

Developing simplified beer statistics. Part 3 – Gravity

 

(Catch up with Part 2 – IBUs, here.)

I want to start by making a note on my use of the BJCP guidelines for this exercise. The purpose is not to necessarily categorize the BJCP guidelines themselves, so much as use the existing guidelines as a list of statistics for “typical beers” to represent what should be pretty much every beer in existence. Basically, I’d like to create a final formula/set of statistical guidelines that can describe any beer. In the absence of an enormous amount of data about existing beers in the marketplace (which would be really preferable), I’m using this as an example set of what the typical range of beers would be if every single beer style were represented evenly in the market. In this case, one of each.

In this post, I’m looking at the average Original Gravity (OG) and Final Gravity (FG) for each style and here’s why:

OG is a measure of how much sugar is present in wort pre-fermentation and FG is a measure of how much sugar is present in wort post-fermentation. Using the two together, you can calculate the amount of alcohol in a beer, which is always a useful statistic. In theory, with these, we can get an accurate sense of how sweet or dry a beer will be. This is, of course, exactly what we need to play against IBUs to attempt to predict what a beer will taste like. Sweeter beers should balance against high IBU beers to create more balance. Dry beers should accentuate hops to give an impression of a more bitter beer.

For the purposes of this I translated OG and FG to “Gravity Units” (GU) by starting with the OG, subtracting 1 and multiplying by 1000. Or, in other words, I just took the last 2 digits of the OG:

1.054 – 1 = .054 * 1000 = 54

Playing with all these numbers has been a little funny. I keep on playing with almost arbitrary math to try to play with these numbers in a way that makes sense to me, and here’s been my stumbling block:

For years, I’ve been using the system laid out in Designing Great Beer by Ray Daniels. It’s somewhat of a biblical text in my house. Sometimes I just read it for fun. In this book, Ray uses a ratio of BU:GU to figure out balance in beer. It’s a good system, but the more I’ve been looking at it while playing with these numbers, the more I think it might be a little – not a lot, but a little – off base.

I sorted my style list by GU – just to get a quick gander of what it looked like. In theory, if the BU:GU ratio is correct, then the beers with the highest GU should be the sweetest and the beers with the lowest should be the driest. And while it’s close, it doesn’t quite work out.

Here’s the list, in order of lowest original gravity to highest.

Berliner Weisse, Scottish Light 60/-, Lite Lager, Mild, Ordinary Bitter, Southern English Brown, Scottish Heavy 70/-, Dry Stout, Special Bitter, Standard American Lager, Blond Ale, Northern English Brown, Brown Porter, Lambic, Scottish Export 90/-, Kolsch, German Pilsner, American Wheat, Weissbier, Witbier, Munich Helles, Cream Ale, Vienna Lager, Schwarzbier, Guezue, Dark American Lager, Dunkelweizen, North German Alt, Bohemian Pilsner, Dusseldorf Alt, Roggenbier, Premium American Lager, Belgian Pale Ale, California Common, Irish Red, Munich Dunkel, Dortmunder Export, Sweet Stout, Classic American Pilsner, Flanders Red, American Brown Ale, American Amber Ale, American Pale Ale, Oktoberfest, Extra Special Bitter, Saison, Oatmeal Stout, Robust Porter, Flanders Brown, English IPA, American Stout, Foreign Extra Stout, American IPA, Traditional Bock, Maibock/Helles Bock, Belgian Dubbel, Belgian Blond Ale, Biere de Garde, Baltic Porter, Old Ale, Weizenbock, Belgian Trippel, Imperial IPA, Belgian Golden Strong Ale, Doppelbock, Belgian Dark Strong Ale, Russian Imperial Stout, Eisbock, Strong Scotch Ale, English Barleywine, American Barleywine

It’s close, but to have things like Lambic, Scottish Export 90/-, and Kolsch all right next to each other doesn’t seem quite right. Same goes for having Southern English Brown, and Scottish Heavy 70/- down on the “dry” end.

If you were sorting in order of alcoholic strength, this would be pretty close.

Just to make sure I wasn’t misrepresenting the stat by taking BU out of it, I also sorted by BU:GU ratio. It’s *also* really close, but you still get things like Strong Scotch Ale (BU:GU of 0.26) and Standard American Lager (BU:GU of 0.256) being directly next to each other. Certainly, the ratio of hops to malt in them are very similar, but the flavor profile of these beers is staggeringly different. On the other end of the scale, I can come up with Ordinary Bitter (BU:GU of .834) and AIPA (BU:GU of .84) right next to each other. Again – ratio similar, flavor profile vastly different.

I keep on doing weird crap to these numbers to try to get something closer. I calculated apparent attenuation (how much sugar has fermented out of the solution) to see if that made sense and that was all over the map. I really wanted to apply attenuation back to the original gravity to see if I could get the high gravity beers with low attenuation to just fall out naturally, but since attenuation is based off of OG and FG I kept coming back to the final gravity. And then I thought something that sounded in my head like, “Duh.”

So I sorted by Final Gravity to see what the list looked like. Here it is from lowest FG to highest FG:

Lite Lager, Guezue, Berliner Weisse, Lambic, Standard American Lager, Flanders Red, Saison, Ordinary Bitter, Dry Stout, Kolsch, Cream Ale, Special Bitter, Witbier, Munich Helles, Dark American Lager, Premium American Lager, Flanders Brown, Mild, Blond Ale, Northern English Brown, German Pilsner, American Wheat, Belgian Golden Strong Ale, Brown Porter, Belgian Trippel, Scottish Light 60/-, Weissbier, Vienna Lager, Dunkelweizen, Roggenbier, Belgian Pale Ale, Irish Red, Biere de Garde, Southern English Brown, Scottish Heavy 70/-, Dusseldorf Alt, North German Alt, California Common, Classic American Pilsner, Dortmunder Export, American Pale Ale,
American Amber Ale, Scottish Export 90/-, Schwarzbier, Munich Dunkel, American Brown Ale, Extra Special Bitter, Belgian Dubbel, Belgian Blond Ale, Oktoberfest, Oatmeal Stout, Robust Porter, English IPA, Foreign Extra Stout, American IPA, Maibock/Helles Bock, Bohemian Pilsner, Imperial IPA, American Stout, Traditional Bock, Belgian Dark Strong Ale, Sweet Stout, Old Ale, Weizenbock, Baltic Porter, Doppelbock, American Barleywine, Russian Imperial Stout, English Barleywine, Eisbock, Strong Scotch Ale

If you were looking for a list of beers, dry-to-sweet, this is pretty damn close. Of course, it makes a lot of sense. The lower the FG is the less sugar is in it. What’s more, because of the limitations of what yeast can actually digest, the higher OG beers will pretty much never ferment out as low as the lower OG beers, with the possible exception of that wacky Saison yeast, so high FG beers will always be sweet, and low FG beers will always be dry, OG be damned.

My next step was, logically, a BU:FU ratio and, again, it’s super close. In order of low-to-high ratio, this would be – in theory – malty-to-hoppy. Clearly, I’m not accounting for pH here, so any sour beers are going to land in the malty end.

Strong Scotch Ale, Lambic, Weissbier, Doppelbock, Eisbock, Dunkelweizen, Weizenbock, Berliner Weisse, Roggenbier, Southern English Brown, Scottish Light 60/-, Dark American Lager, Scottish Heavy 70/-, Traditional Bock, Baltic Porter, Witbier, Belgian Dubbel, Belgian Dark Strong Ale, Standard American Lager, Guezue, Mild, Sweet Stout, Oktoberfest, Belgian Blond Ale, Scottish Export 90/-, Munich Dunkel, Irish Red, Munich Helles, Biere de Garde, Cream Ale, Maibock/Helles Bock, Premium American Lager, Vienna Lager, Blond Ale, Schwarzbier, Belgian Pale Ale, Dortmunder Export, American Wheat, English Barleywine, Flanders Brown, American Brown Ale, Oatmeal Stout, Northern English Brown, Brown Porter, Old Ale, Flanders Red, American Amber Ale, Classic American Pilsner, North German Alt, Bohemian Pilsner, Robust Porter, Belgian Golden Strong Ale, Belgian Trippel, Kolsch, Russian Imperial Stout, American Pale Ale, California Common, Extra Special Bitter, Special Bitter, Lite Lager, German Pilsner, Ordinary Bitter, Dusseldorf Alt, American Stout, English IPA, Foreign Extra Stout, American Barleywine, Saison, American IPA, Dry Stout, Imperial IPA

Nice list! The only real problem is that, in terms of creating a statistic, our numbers start at .702703 (Strong Scotch Ale) and end at 6 (IIPA). It’s a really weird metric – it’s good for making this list, but not for saying something like, “This beer is a 35? Woooo! Hoppy!” Which is, essentially, my end goal here.

So next up? Making these numbers into a form that’s easier to digest than .702703. With any luck, this should be pretty quick.

In the meantime, homebrewers (or hey.. commercial brewers if you guys are reading this with baited breath), can you do me a favor?

Go do a BU:FU calculation on some of the beers you have logged into your brewing journal. Get the end number, (like .702703), multiply by 20 (it’s an adjustment), then tell me the style of beer, the number you came up with, and if YOU think the beer was malty, hoppy, or balanced.

It goes like this:

Let’s say you made a Vienna Lager, FG of 1.012, 24 IBUs.

24/12 = 2
2 * 20 = 40

“Erik, I made a Vienna Lager, the number is 40, I thought it was quite well balanced.”

Tags Tags: , , ,
Categories: appreciation, industry, marketing, op-ed
Posted By: erik
Last Edit: 13 Nov 2009 @ 03 12 PM

EmailPermalink
 

Responses to this post » (3 Total)

 
  1. christopher says:

    I’ve only got IBU calculated on a few beers, if this method proves compelling I can go back and get you more data.

    My Oatmeal Porter scored 23.6 (26IBU:1.022FG) and is fairly balanced but tipped towards malty. Its thick and heavy.

    Vanilla Brown Porter scored 27.5 (22IBU:1.016FG) and was more malty and sweet tasting despite the lower FG.

  2. BeerBarn says:

    Well this is good information to know when I start making my own home brewed beer. I will post the info. once I make my first batch.

  3. erik says:

    Awesome – I appreciate it. I really have to hound more people for stats so that I can compile some useful data. At home my beers match my findings, but I would hardly consider myself a good test case for my own math.

Post a Comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

\/ More Options ...
Change Theme...
  • Users » 128712
  • Posts/Pages » 204
  • Comments » 2,670
Change Theme...
  • HopsHops « Default
  • BarleyBarley

About



    No Child Pages.

Shirts



    No Child Pages.

Tour



    No Child Pages.