The Trouble With Beer Numbers

A friend who was dining at a McMenamins pub recently sent me a photo of the beer menu with a technical question. (It wasn't reproducing well, so I grabbed the picture at the top of the post off the internet.) To their great credit, the McMenamins have finally started to take beer seriously. They broadcast this to their customers by offering detailed descriptions of their beers followed by a laundry list of stats. However, I think the storm of information may be more confusing than explanatory. Here's what they tell you about Nebraska Bitter:

Malts: Premium 2-Row, Crystal 40
Hops: Cascade
OG: 1.044 TG: 1.009 ABV 4.52% IBU: 20 SRM: 4

Before I go any further, let me show you a second, representative example of what you find on a typical wine menu:

With wine, you get the grape, the place of origin, and the price. Sometimes you get a vintage (but too rarely). Whisky lists are much the same--a type (rye, Irish, etc), possibly the length of aging, and a price.. These may or may not be a good systems; particularly with wine, you're at the mercy of your own knowledge and the helpfulness of your server. But at least you're not offered a bunch of potentially confounding information.

What do all these details tell us? At the best, they provide accurate information to the small minority of people who know what they mean. At the worst, they provide misinformation. Mostly, though, I think people just tune them out. You really have to know a lot about beer to interpret them and, even then, they are for the most part not that revealing. Let's dive a little deeper:

Malt and hop varieties. There are some cases when an ingredient is noteworthy. Mostly, this information is of limited value because we have both too little and too much information. What proportion of the grist do the various malts comprise? ("Premium" tow-row? God knows.) The same is true of hops. Seeing a list of several varieties tells you basically nothing unless you hate one of them. When they were used in the process, how many were used, and what percent of the total amount--without knowing this info, the hop composition is basically useless.
Gravity. No one knows what this is unless they brew. "Gravity" is a specialized term that indicates the amount of sugars in solution (relative to water). The difference between OG ("original gravity," or the measure before fermentation) and TG or FG ("terminal" or "final" gravity, the measure after fermentation) tells you how much of those sugars were converted. There's nothing in these terms that you can't discern from looking at the alcohol content and assessing the mouthfeel for attenuation.
Alcohol by volume (ABV). Actually very useful!
International Bitterness Unit (IBU). This is perhaps the most problematic of all the numbers. It's supposed to tell you how many bittering compounds are in a beer, but the number (a) is wildly variable, and (b) cannot tell you how bitter a beer will actually taste. The worst thing about IBUs as a metric is that they're not even precise. If a brewery uses an algorithm to calculate this number based on the reported hop acid levels, the result will deviate--always, to some degree--from the figure derived from chemical analysis. How widely can these figures vary? Massively; by a 100% in some cases. Moreover, we now know there are lots of bittering compounds in hops, and the IBU figure may or may not take that into account. Nearly as problematic is the little-understood relative nature of IBUs. They can be balanced by other elements in a beer, particularly sugars, so a barleywine with 40 IBUs will taste very sweet, whereas a session IPA with 40 IBUs will be sharply bitter.
Standard Reference Method (SRM). This is supposedly an indication of color, but in order to be precisely-calculated, it has to be measured in lab conditions: with a particular vessel under precise lighting. You can't very well gauge it by comparing a flute of beer with a goblet. And since the beer comes in, you know, a transparent glass, one can assess that color fairly easily without resorting to number. (Let's see, is 10 SRM golden or copper...?)

How Should We Talk About Beer?
Beer comes in so many varieties (styles, colors, strengths, dominant flavors, etc.) that communication is a critical element. One of the biggest barriers breweries have introducing new beers, particularly in offbeat styles, is figuring out how to talk about them. There's a big need here.

One approach would be purely descriptive. This is certainly one important element of the communication, but it has its own flaws. Here's a commercial description of a beer. Tell me if you can figure what kind of beer it refers to.

“It has a dense head and a wonderfully spicy aroma with notes of dried fruit. Full-flavored yet also effervescent and refreshing, the flavor is nutty and caramel-like but with a dry, lingering finish that is gently hoppy.”

Scottish ale? Marzen? Dubbel? Actually, that's La Choulette's description of their Ambrée bière de garde. Descriptions, particularly when they're meant to move product, often slide into a word salad of adjectives, using specific words like "lemony" to describe subtle tonal elements or generic terms like "hoppy" or "malty" to describe dominant ones.

I've been pondering this for awhile, and I've come to believe descriptions are best--but with some structure. There are a few categories of information we care about: the dominant flavors, the style inspiration, key ingredients, body and carbonation, and overall flavor impact. That last one is one of the most important, and one we haven't ever really addressed. If I describe a helles and an imperial stout, the adjectives might not tell you that the latter has a towering intensity while the former barely moves the needle. To be useful, descriptions need to walk through the same categories of information each time.

How are these men describing their beers? — How are *these* men describing their beers?

Almost as a thought experiment, I came up with these descriptions, which blend the abstract and concrete. It allows you to compare beers on the menu with each other, apples to apples. Have a look:

Alworth's Best (4.1%)
A malt-driven beer
Alworth's Best is our intepretation of an English pub ale. It uses English malts to create slightly sweet, richly bready flavors (think scone). A fruity note comes from the yeast strain we use, and this is enhanced by Goldings hops, which add their own marmalade signature. It has a light body and light effervescence. It's a great beer to drink in clusters of twos and threes.
Flavor intensity: moderate-low

Spring Sunset (5.5%)
A yeast-driven beer
Spring Sunset is our interpretation of the 19th century ales made in the farmlands of Belgium and France. We included wheat and oats to give it a grainy, breakfast-cereal flavor and a touch of haze in the glass. Most of the flavor comes from a yeast strain famous for creating peppery and herbal flavors along with bright lemongrass fruity notes. Most of the flavor and aroma in this beer comes from the yeast. We finished it off with French hops, which add a tiny bit of bitterness and a wildflower aroma. It is light-bodied and highly effervescent.
Flavor intensity: moderate

Portland Pride (6.2%)
A hop-driven beer
Portland Pride is our effort to create the perfect Oregon IPA. We used a touch of honey malt to give the beer some sweetness, and then infused it with four varieties of hops to create a rich, aromatic ale. Nearly all of the flavors and aromas of Portland Pride come from the hops, and we gave it a moderate level of bitterness for balance. The very floral nose comes from the Cascade hops we used to dry-hop the beer. The flavors are fruitier (we taste peach and white wine grape) and more citrusy (lemon and orange); these come from the Amarillo and Galaxy hops. It has a medium body and light carbonation.
Flavor intensity: high

I have no doubt they could be improved upon. In fact, I think it would be great to hear ideas. The biggest challenge--and one I'm not sure I have cleared--is characterizing hops. Reducing them to bitterness is not adequate, but that's an important detail. IBUs are fatally flawed as a measure (even when you know how to interpret it, you can't be sure the brewery measured it accurately), but I would like a sense of how bitter a beer is going to taste. With the dominance of IPAs in the US, we also need a sense of the intensity of flavor and aroma, too. How "juicy" is the beer? That's a special area of improvement I need to tackle.

I do think the numbers game is a blind alley. Except for ABV, none seem to get us any closer to understanding a beer, and some actually confuse matters. So I leave the final word to you. I've given you my best rough-draft effort. If we dump numbers and start from scratch, how would you create a uniform system of beer descriptions?