Wednesday, April 15, 2009

Root Beer Floats: A Quest for Perfection, Part I

It occurred to me recently that in several years of penning a blog with "root beer" in the title that I haven't paid due attention to the pinnacle, the king, the gateway drug of the root beer world: the root beer float. Most people that I know who have made the transition from root beer hater to root beer lover have been converted by the silent foamy proselytism of the root beer float.

When done properly, the root beer float (hereafter RBF) is a magical thing, far greater than the sum of its parts. However, when done poorly, an RBF is a sad, unfortunate thing, so I think it's time to deconstruct the RBF and analyze exactly what goes into making the perfect RBF. This isn't a hopeless quest: unlike many foods, e.g., pizza, RBFs can actually be made better by you at home than you will ever find in a restaurant. It’s certainly possible for a restaurant to make an awesome RBF, but 99 times out of 100, they either don’t care enough to make a good one, or simply don’t know what it takes to make the perfect RBF.

Let's consider the variables:
  1. The Ice Cream
  2. The Root Beer
  3. The Ratio
  4. The Vessel & Utensils
  5. The Process
  6. The Ineffable
In Part I, I will be addressing 1. and 2., the two most essential components of any RBF.

1. The Ice Cream

First and foremost, the vanilla ice cream used in the perfect RBF must have a true vanilla flavor. Ice creams using synthetic vanillin or other artificial flavors are not going to cut it. A lot of people, including me, grew up enjoying the artificial stuff so I admit that it can have a nostalgic element to it, and I appreciate that this borders on putting lipstick on a pig, but the flavor is simply far superior when you use a real vanilla ice cream. Whether the ice cream uses natural vanilla flavoring and/or includes vanilla seeds (e.g., Häagen-Dazs’s “Vanilla Bean” or Ben & Jerry’s Vanilla) doesn’t really matter much as long as the flavor is good and not cloyingly sweet. Expensive gelatos are not necessary, in fact you don’t want something so rich that you only want a few bites or so dense that you can't get the spoon into the ice cream when it's bobbing around inside the float.

Try to reconcile A, B, & C with D & E

I picked up a package of a cheap boxed ice cream as an example of an ice cream that you don't want to use for an RBF (and also for the video experiment below). Incidentally, this is another example of why you have to read food labels closely, especially in the US where blatant trickery is the norm. This package is simultaneously implying that they grow vanilla orchids and/or raise cows on a farm (presumably with a bay view) to produce "plain vanilla goodness" for this ice cream, yet the ice cream is artificially flavored and contains no vanilla whatsoever. Essentially they're trying to trick you into thinking you're getting something good and hoping desperately that you won't notice that it's mostly air, fillers, and, on top of that, it really doesn't taste very good.

The second major factor is overrun. Overrun is the percentage of air mixed into ice cream, calculated as the percentage increase in volume of the ice cream relative to the original liquid mixture. The legal overrun limit in the US is 100%, or an ice cream that is half air. Bayview Farms and other economy brands push that limit as far as they can, regularly having overruns in excess of 90%. Gourmet gelatos commonly will have ~20% overrun, whereas many of the major ice cream brands will have overruns ranging from 40–60%. Nothing is worse for an RBF than an ice cream that melts into the root beer too quickly turning the RBF into a creamy mess with no distinction between the liquid and the solid, so avoid high overrun ice creams.

To illustrate the effect of overrun, I did a simple experiment. I placed equal amounts (by volume) of a cheap high overrun ice cream (the Bayview Farms pictured above, with a density of 0.54 g/ml) and a common low overrun ice cream (Häagen-Dazs Vanilla , with a density of 0.90 g/ml) in two room temperature bowls side by side and simply watched them melt. Here is a little video of the results for your enjoyment:

Bayview Farms (left) vs. Häagen-Dazs (right)

I realize that this was a waste of ice cream, but sometimes you have to be cold as ice and willing to sacrifice for science. Not surprisingly, the Bayview Farms started visibly melting immediately and liquefied well before the Häagen-Dazs. Given that a scoop of any ice cream put on a plate next to a piece of pie will seem to melt into a puddle within seconds, I was surprised by how long it took for the two ice creams to fully liquefy: 1 hr 25 min for the Bayview Farms vs. 2 hr 51 min for the Häagen-Dazs.

Thirdly, the ice cream must be fully frozen before being added to the RBF serving vessel; allowing the ice cream to soften enough to scoop is fine and actually desirable, as the ice cream should not be too hard to eat in the RBF.

There are a few other factors to consider when selecting an ice cream. The fat content of ice creams ranges from ~10%–20%, and, predictably, the more expensive ice creams tend to have higher fat contents (as well as low overruns). Some brands make custard-style ice creams that include egg, sometimes called “frozen custard” – some brands advertise their egg content (e.g., Ritter’s Frozen Custard), others don’t (e.g., Häagen-Dazs and Ben & Jerry’s). In general you can ignore these factors — flavor and overrun trump all others: if you get a tasty, low overrun ice cream, you're set.

I do have to say one word in defense of soft-serve ice cream: it's artificial, it's high overrun, it's questionably ice cream at all, so it breaks all of the above rules, yet somehow it can really work in an RBF. They're not the perfect RBF, they're ridiculously sweet, and the soft-serve melts into the soda very quickly, but I'd never turn up my nose at an RBF from an A&W restaurant.

2. The Root Beer

The root beer choice is the most difficult and contentious element of making the perfect RBF. Despite the obvious importance of using the right root beer, most restaurants don’t give this due consideration, generally opting to serve whatever they have in the soda dispenser. While I admit that the classic A&W RBF has a place in my heart, A&W root beer is really not a good choice for making the perfect RBF because it is too sweet to start with, and it has a strong vanilla note, so when you add vanilla ice cream you will get an overly sweet and vanilla-y float. The goal is to find a root beer and ice cream pair that complement each other. The upshot of this is that some of the best root beers for drinking plain will not be good choices for making the perfect RBF.

A variety of root beers mapped on a two-axis root beer space: dry to sweet, and vanilla to herbal (i.e., predominantly wintergreen, anise, or other spice notes) detected with highly sensitive rootbeerometers. The "sweet spot" for RBF-making lies within the dashed line. X= root beers that are no longer extant. The bottled A&W Float broke the rootbeerometers, so its location in root beer space is an estimate.

The optimal characteristics for an RBF root beer include: not being excessively sweet, a moderately high degree of carbonation, and a balanced flavor that leans towards the herbal side of the spectrum (i.e., not overly vanilla-flavored). To break it down one step further, root beers with prominent wintergreen/birch notes are often a good choice as opposed to those with anise/licorice notes. I’m not a zealous proponent of sugar over corn syrup in general, but in this case the sugar-sweetened root beers tend to be drier, spicier, and have a higher freezing temperature that promotes the formation of ice crystals for “frozen edges” (discussed in detail in Part II). The root beers that best fit the above criteria, are shown in the "sweet spot" in the above graph. Barq's is by far the most common of the root beers that fall in the sweet spot, but it should be noted that this is the one root beer on the graph that contains caffeine.

Barq's: a pretty good choice for RBFs, but it does contain caffeine

Because of my stand against artificial flavors for the ice cream, one might expect the same rationale to apply for root beer. However, most all-natural root beers (although not all) fall short in terms of flavor, mostly because the use of sassafras in food and beverages has been illegal since 1960 in the US due to the toxicity of safrole. There is no substitute for sassafras apart from artificial flavoring agents which can mimic (to a degree) the sassafras flavor, so often times I find myself preferring those that have a small amount of artificial flavor to supplement the natural flavors. So sue me.

Safrole: it may be carcinogenic, and it may be a precursur to MDMA, but it's mighty tasty

With the exception of Abita, all of the root beers in the RBF sweet spot contain some artificial flavors. If you're totally opposed to artificial flavors, the best choices are Abita, Virgil's, or Blue Sky.

While this may seem to be all you need to know about making RBFs, there are more factors to consider before true perfection can be attained. Stay tuned for the next post where I'll cover the remaining variables that go into making the perfect RBF.