Last time around I talked about American stouts and the relationship between quality and alcohol content. There were two clear conclusions from that: one, stronger stouts are generally better than weaker ones, and two, almost no stouts under 6% ABV are worth your time and money. News to use, right?
Of course, American stouts are just one of the many stout styles out there. While they might be the most commonly available options, it'd be unfair to draw conclusions about the whole stout family based on one style. What happens when we expand our data set to look at all stouts? Take a look at this chart, showing ABV% and BAR for 897 stouts.
OK. First off, that one dot is way the hell out there. That would be BrewDog's Tactical Nuclear Penguin. It's 32% ABV, and as far as I can tell it's more similar to Jägermeister than it is to most other stouts. They get it to 32% by freezing it and removing as much water as possible, which you may recognize as the same process used to make Natural Ice from Natural Light. It's served in shot glasses. Given all that, it'd be a bit silly to compare Tactical Nuclear Penguin to other stouts. Let's drop that point and see if the chart is a little easier to make sense of.
A little better, but there's still a lot happening there. Perhaps this will make more sense if we look at it in buckets, rather than trying to include 900 data points on one chart?
This jibes with what we saw with American stouts, for the most part. Very low ABV% stouts still don't look to be a great buy, though it's worth noting that there are apparently some decent stouts in the 5-6% range. Those tend to be oatmeal, milk, and cream stouts, by the way; you won't find many exceptional offerings, but those styles will be your best bets if you're looking for a lower-alcohol stout.
If you're just interested in getting the best stouts overall, it's pretty clear that you want something that is 10% ABV at the very least. Your chances of getting a great beer improve even more if you pick something 12% or above; there's a pretty clear jump at each percentage point between 10% and 12%, after which the ratings mostly level off.
There are spikes at the 14-15% and 19-20% ranges and a dip at the 15-16% level, but I wouldn't read too much into those. Going into this, I'd been expecting that there might be a point where the alcohol was just too much; at first, it seems like we're seeing that at around 15%.
However, that 15-16% bucket includes a total of only five beers, including the only >13% ABV stout to score a negative BAR: He'Brew Jewbelation Sweet 16. Seems unfair to judge the other four stouts for that one's poor performance, no? Removing that beer from the group would raise that average to just over 4.5 BAR, so I don't think there are any conclusions to be drawn there.
Just as the average in our 15-16% range suffers from one bad apple, the spike at 14-15% is driven almost entirely by one amazingly excellent apple: Goose Island's Bourbon County Brand Stout. There are several iterations of this particular brew, but three of them rate at 10 BAR or above, and all three of those are between 14 and 15% ABV. That's a quarter of the sample in that range, so the only real information we get from that is that Bourbon County Brand Stout is phenomenal in all of its versions. If you take just one thing from this article, let it be that you should strive to acquire some Bourbon County Brand Stout as soon as possible. Trust me on this one.
Anyway, given how much noise we're seeing from the small sample sizes here, maybe it makes more sense to just put everything above 12% in one bucket. Take a gander:
Within this >12% ABV group, we're talking about a group of 69 stouts that average 4.84 BAR. Just for comparison, there are 223 stouts at 6% ABV or less; exactly two of those surpass 4.84 BAR. Nearly half of those low-ABV stouts score negative on the BAR scale.
On the other hand, just two of those 69 super-high-ABV stouts score a negative BAR. 38 have a BAR above 4. Basically, you could buy a strong stout completely at random and you'd have good odds to get a well-above-replacement level beer. Trying that with randomly selected low-ABV stouts is unlikely to satisfy.
Now, let's go back to that chart with all of our stouts, because there's a couple unique cases that are worth discussing. First, that He'Brew Jewbilation Sweet 16 I mentioned seems like it might be pretty bad, right? Take a look here; it's the point highlighted in red.
There's really no comparable beer; Sweet 16 is the only very strong stout that is rated so poorly relative to its style. However, imperial stouts have a much higher replacement level than other stouts. Sweet 16 has an average Untappd rating of 3.27. For imperial stouts, that translates to a -.18 wOBAR. If it were classified as just a standard stout, it'd have a wOBAR of .15 instead, and it'd be approximately a 2 BAR beer. Considering that, I'd have to say that the high scores for other imperial stouts are all the more impressive -- not only are they far and away the best beers by BAR, but the style they're competing within has one of the highest replacement levels on our site.
You may have noticed two other points that I highlighted in that last chart; the green dot is Guinness Draught, which is an Irish Dry Stout coming in at .71 wOBAR and 9.68 BAR. There's reason to doubt this rating. It has over 21,000 checkins, but 3,000 come from one person, who checks in as many as fifteen times a day with a 5-star rating for Guinness. I'll leave you to make your own conclusions there, but I think it's safe to say we should dock Guinness some points for that.
Even if we accepted Guinness' rating as legitimate, though, there's reason to look closer at how it ended up so high. See that purple dot in the center of the chart? That's New Holland's Dragon's Milk. It's an imperial stout with a wOBAR of .19, so at first glance, we'd say that it's not only less popular than Guinness, but also just not as well-liked a beer. However, if you look at just the raw Untappd scores, Dragon's Milk has an average of 3.64 stars. Guinness Draught? Also 3.64.
Guinness' high BAR benefits from its popularity, sure. That said, it's also getting a big boost from being a less common style. Irish dry stouts just aren't great on the whole, so Guinness ends up being a big fish in a small pond, whereas Dragon's Milk struggles to separate itself from the masses of great imperial stouts.
Obviously, that's part of the idea behind BAR; we give beers a boost if their style is one of the lower-rated ones. Still, it's sometimes worth taking a step back. If you went purely by BAR and didn't consider the components of that score, you'd conclude that Guinness Draught is just as good as Goose Island Bourbon County Brand Stout, and that would be a tragic error.
You can follow Alex on Twitter @AlexanderFossi.