Dilution Is Not a Percentage. It Is a Function of What the Drink Already Is.

The reason a single dilution number never works is that dilution is not a property of the procedure. It is a property of the drink. Two drinks stirred for exactly the same twenty seconds, with the same ice, in the same glass, want different amounts of water, and if you use one number for both you will get one good drink and one mistake.

What decides it is mostly arithmetic, and the arithmetic is simpler than the way it usually gets discussed.

The main variable is how far the base is above where you want to land

The primary job of dilution in a cocktail is to take a mixture that is too strong to drink and bring it down to a strength you want to drink. Everything else dilution does is secondary to that.

So work out where you are starting. Take every ingredient, multiply its volume by its alcohol percentage, add them up, and divide by the total volume. A Martini at five parts gin to one part dry vermouth, with a gin at forty-seven percent and a vermouth at seventeen, starts at about forty-two percent alcohol. An Old Fashioned built with two ounces of a forty-five percent bourbon and a quarter ounce of syrup starts at about forty. A Daiquiri with two ounces of rum, an ounce of lime and three quarters of an ounce of syrup starts at about twenty-one, because the juice and the syrup are already mostly water.

Then decide where you want to end. Spirit-forward drinks land well somewhere around twenty-eight to thirty percent. Sours land well around fifteen to seventeen. Those are preferences, not laws, and they are the numbers worth arguing about rather than the dilution figure itself.

The water you need is roughly the starting strength divided by the target, minus one. Forty-two divided by twenty-nine is about one point four five, so the Martini wants about forty-five percent of its volume in added water. The Old Fashioned at forty over thirty wants about thirty-three percent. The Daiquiri at twenty-one over sixteen wants about thirty percent, which is what a hard shake delivers.

There is the factor of two, and it is not a contradiction between sources. It is the difference between a Martini and a Daiquiri.

The second variable is how long the drink is going to live

A drink served up in a coupe is finished the moment it is poured. Whatever water is in it is all the water it will ever have, so it has to arrive complete.

A drink served over ice is still being made while it is being drunk. It picks up another five to ten percent in the first few minutes and keeps going, and a large clear cube slows that down but does not stop it. Which means a drink on the rocks should leave the mixing glass deliberately under-diluted and slightly too strong, and it should taste correct at the midpoint of the drink rather than at the first sip.

This is the most commonly missed adjustment in home bartending. People taste the first sip, find it sharp, add water, and then spend twenty minutes with a drink that is going quietly flat.

The third variable is whether something else is going to dilute it later

If there is a sparkling component going on top, that component is the dilution and there is no other dilution budget. A highball built with three ounces of soda over a two-ounce base does not want any pre-dilution at all, and a stirred-down base under soda is a drink that tastes correct for exactly one sip.

The same is true of anything served long over a lot of ice on a hot day, and it is true in the extreme of anything with no alcohol in it. A zero-proof batch has nothing to bring down, so every drop of water is pure loss. That arithmetic runs differently end to end.

Sugar and acid change how much dilution the drink can absorb

Strength tells you how much water the drink needs. Sugar and acid tell you how much it can take before it reads as watery, which is a different question with a different answer.

Dissolved sugar adds viscosity, and viscosity is most of what the mouth uses to decide whether a drink is thin. A sweeter drink tolerates more water before it goes slack. A bone-dry one has nothing holding it up and falls apart fast, which is why a very dry Martini is so much less forgiving of a long stir than a Manhattan is.

Acid behaves differently again. Dilution drops perceived sourness roughly in proportion, so an over-diluted sour does not taste watery so much as it tastes dull and slightly sweet, because the sugar is still registering and the acid has fallen below the line where it was doing its work.

Tannin is the quiet one. A drink carrying real tannin, from tea or a wine-based component, holds its shape under dilution far better than anything else, because astringency is a textural signal and it survives being watered in a way that flavor does not.

Which is why the rule should be a question, not a number

Before you stir anything, ask three things. How far above drinking strength am I starting. How long will this drink sit in the glass. Is anything else going in later.

The answers give you a range, your own measured stir gives you the actual figure, and the two together mean you never have to trust a percentage from somebody else’s kitchen again.

And if the drink is going in a glass for a party rather than a glass for one person, the answer changes again, because batching and building are two different decisions before they are two different techniques.

Part of Drinks & Zero-Proof.