Ben LaVerriere

CREATE VIEW what_to_drink

SQL source code creating a view called "what to drink"

Years ago I received as a gift Vintage Spirits and Forgotton Cocktails, by Ted Haigh (alias Dr. Cocktail). I’ve made a smattering of drinks from it by now — and happened across more in the wild — but one question stuck in my mind from the beginning:

If I’m going to invest in some of the wackier ingredients, which ones will let me make the most drinks?

This sounded like a database problem. One GUI, one normalized-enough™ schema design, and a lot of manual data entry later, I have my answers. I won’t be sharing most of the details here — buy the book and/or come visit me! — but I will reveal a few tidbits.

By appearance count, the top five ingredients are lemon juice, gin, Angostura bitters, lime juice, and orange juice. Somehow I expected either gin or lime to take the lead, and I wouldn’t have guessed that orange would be anywhere near the top given the cocktails I’ve come across. Something new to explore!

The top five most frequent liquors are gin, orange curaçao, brandy, dry vermouth, and sweet vermouth. (This isn’t counting a few specific vermouths or similar aperitifs that are mentioned by name or character, like “dry French vermouth”.)

By approximate total volume, gin tops the list at about 40.5 ounces, followed by bourbon at 15.5 (even though bourbon only shows up in six cocktails). This probably has a lot to do with the period that birthed many of these recipes — but it’s interesting to note the trend towards heavy pours of bourbon when it does show up.

There are 55 ingredients (!) that only appear once, from kümmel to falernum and from Carpano Antica to Amaro Cora. But that’s only if we differentiate…

…the twelve different kinds of rum specified throughout the book. (This might already represent a simplification, with some coalescing that happened as I transcribed ingredients.) There’s some interesting data-model questions here, about “valid” ways to combine different types of rum without screwing up the individual recipes.

And finally, returning to the rare and unusual ingredients that started this whole project, the top five ingredients I’ve never used in my home bar: real pomegranate grenadine, maraschino liqueur, Calvados, apricot brandy, and raspberry syrup! Some of these are already favorites of mine to drink, so I’m excited to get some mileage out of them. (Backbar taught me, via the Hotel Nacional Special, that I adore apricot brandy in ice-cold drinks.)

Every bit of this project is slightly imprecise. But at the end of the day, I’ve now got a tiny electronic brain I can consult, and this brain knows how to answer exactly the kind of questions I wanted.

Coming up next chez nous: the Barbara West cocktail, because we are a household that loves sherry. (Sherry is a singleton ingredient but it’s counterbalanced by that popular trio, gin + lemon + Ango.) Cheers!

Source code prettification by Carbon.