How classic cocktails are actually built

1. Classic sours use less sugar than sourness

Across 36 short sours, the median uses 0.37 parts citrus to each part of spirit, and 0.75 parts sweetener to each part of citrus. Against 60ml of spirit that is about 22.5ml of citrus and 17.5ml of sweetener — sweet under sour, not the equal measures of the often-quoted 2:1:1.

What sweetens the drink matters. Sours sweetened with a syrup (27 of them) sit at a median of 0.75; those sweetened with a liqueur (9) at 1.00. A liqueur is less sugary than syrup and brings alcohol with it, so it is poured more generously — which is why a Margarita and a Sidecar run equal parts of liqueur and citrus while a Daiquiri does not.

Population. 73 published drinks have a sour structure. Included: the 36 that are short (nothing lengthening them), soured only by fresh citrus, and have a measured spirit and sweetener. “Spirit” includes alcoholic modifiers such as vermouth. Ratios ignore how sweet each syrup or liqueur is and how acidic each citrus is; rich syrups and honey are counted by volume like any other. The interquartile range is 0.67–1.00 for sweetener to citrus and 0.29–0.42 for citrus to spirit. Try a ratio yourself in the Spec Lab.

2. Structure predicts strength better than name

A drink's structure — not its base spirit — decides most of how strong it is. The middle half of each structure, after the water ice adds:

Method. Strength is computed from each recipe's measures and typical bottled strengths, with the method's typical dilution from the site's dilution ranges. Structures with fewer than eight recipes are left out. A spirit-forward drink is small and strong; a long drink is larger and lighter, though the units in a glass can be similar.

3. Alcohol-free drinks are built long

60% of the 40 alcohol-free drinks on CocktailHQ are long, lengthened drinks, against 27% of the 153 that contain alcohol. Short, concentrated structures — spirit and modifier, spirit and sugar, bitter aperitifs — almost disappear without alcohol.

That is not a gap in the catalogue so much as a fact about alcohol. In a short drink, the spirit supplies weight, warmth and a bitter edge; take it out and what remains is a small glass of sweetened juice. Length, carbonation and bitterness from tonic, tea or grapefruit are how alcohol-free drinks get that structure back — the alcohol-free guides explain how.

Population. All published recipes. Low-alcohol drinks are counted with those containing alcohol. Structure is read from each recipe's ingredients by the rules on every recipe page.

4. Three glasses serve 70% of drinks

A coupe glass, a highball glass and a rocks glass between them serve 135 of the 193 recipes on CocktailHQ. It takes 7 glass types to reach nine in ten; the rest of the cupboard — tiki mugs, copper mugs, punch cups — exists for a handful of drinks each.

For a new home bar that is a buying order. A highball and a short tumbler cover the long and the stirred; a coupe takes everything shaken and served up. The fourth, a wine glass, adds only 11 more. Each recipe's Glass and ice panel says what size it needs.

Method. All published recipes. A recipe counts as served by its own glass or by an alternative serve the recipe itself records, never by a glass it merely fits. Each step adds the glass that serves the most recipes not yet served; ties go alphabetically.

5. Most drinks can be at least partly made ahead

Of 193 recipes, 38 batch completely, 128 can have their stable base mixed ahead with something fresh added as you pour, and only 27 (14%) have to be made from scratch each time.

Structure decides it. Spirit-and-modifier drinks — Martinis, Manhattans, Negronis — have nothing in them that spoils, so most batch whole. Sours almost never do: fresh citrus fades within hours, so the spirit and sugar go in the bottle and the juice goes in at service. The drinks that must be made to order are mostly built on crushed ice — juleps, swizzles, tiki drinks — with the rest blended or topped with a fresh shaken foam. Even those can be measured out in advance. How to batch.

Method. All published recipes, each with the batching verdict its own recipe page shows. That verdict comes from how perishable each ingredient is and whether the drink needs a shake, a foam or a fizzy top at service.

6. What does the work in an alcohol-free drink

Take the alcohol out and something else has to supply the structure. Across 40 alcohol-free drinks and 153 with alcohol, acidity is just as common either way. Bitterness is not: 20% of alcohol-free drinks have an ingredient that brings it, against 25% of cocktails — and where it appears, it comes from alcohol-free aperitif, black tea, strong-brewed, espresso, tonic water. Aroma runs the other way (25% against 13%): herbs and spice stand in for what a spirit would have smelled of.

That gap is why so many alcohol-free drinks taste like juice. The guide to bitterness without alcohol covers the fix.

Method. All published recipes; low-alcohol drinks count with those containing alcohol. Each job is read from the ingredient roles shown on every recipe page. Bitterness also counts any ingredient scoring 3 or more out of 5 for bitterness, because a bitter aperitif is often the base of a drink rather than an accent in it.

7. A cocktail’s alcohol is not all in its spirit

In 43 of 134 drinks, a quarter or more of the alcohol is in something other than the base spirit — the vermouth, the liqueur, the aperitif. Among spirit-and-modifier drinks it is 15 of 24. 51% of a Negroni’s alcohol is in its Campari and vermouth, and 69% of a Last Word’s is in its liqueurs; a Daiquiri’s is all rum.

That changes how to make a drink lighter or alcohol-free. Swapping only the spirit leaves a stirred drink with much of its alcohol, which is why the Alcohol-Free Negroni replaces all three of its parts, and why cutting the base measure lightens a Negroni less than it lightens a sour. Lighter drinks shows the reduction for each recipe.

Method. Published recipes containing alcohol, low-alcohol included. Pure alcohol for each ingredient is its measure times its typical bottled strength; dashes of bitters are left out, as they are from every units figure on the site. Base and modifier are the roles shown on each recipe page. The 19 drinks built on two or more spirits are left out, because which of them counts as the base is a matter of labelling rather than of the drink.