Words by: Simon Difford
I’m guessing you're reading this page because you’re already an accomplished cocktail mixer and are interested in the fineries of the art or you've just read Cocktail shaking basics and were intrigued to see what more there is to know. Either way there’s more to shaking than you’d think.
I've been mixing up cocktails for well over 20 years and during that time little changed, until 2009 when Dave Arnold, then of New York's French Culinary Institute, and Eben Klemm, a much respected bartending friend of his, presented The Science of Shaking at that year's Tales of the Cocktail. Dave expands on this and his subsequent research in his excellent book Liquid Intelligence and much of the following is drawn from his work.
Stainless steel heats up and cools down quickly, and in doing so uses minimal energy, so having little effect on the temperature of the finished cocktail. In contrast, heavy boston glasses have more thermal mass, so absorb more energy from the drink being shaken. So steer clear of glass and use all stainless steel 'tin and tin' shakers.
I've noticed that most bartenders assemble their cocktail in the smaller tin of a two-piece shaker and then also scoop their ice into this smaller tin. This risks the drink splashing the ice scoop so contaminating the ice and limits the amount of ice going into the shaker compared to scooping the ice into the larger tin.
Ideally combine the ingredients in the smaller tin and then add ice to the larger tin. If you use one quick motion you can bring the tin filled with ice down on the smaller tin without losing any ice or spilling your drink. However, to play safe pour the contents of the small tin into the large ice-filled tin before sealing.
Dave Arnold's experiments have proven that no matter how you shake and pretty much whatever kind of ice you use, as long as you shake vigorously for 12 seconds, you'll produce a drink between -5°C/23°F and -8°C/18°F (differences mostly dependent on construction of shaker, its temperature and alcohol strength of cocktail) with very much the same dilution. Dance around as much as you like but different shaking techniques make no difference to the end temperature or dilution. And that includes the 'Japanese hard shake'.
The effect of a large amount of ice being forced back and forth through a relatively small amount of liquid in a cocktail shaker is such an efficient way of cooling that after around 12 seconds thermal equilibrium is reached, between -5°C/23°F and -8°C/18°F (more alcoholic drinks can be colder due alcohol freezing at a lower temperature than water). Continuing to shake after this point is reached will make very little difference to the temperature or dilution of the drink.
Ice taken straight from a freezer is usually around -18°C/-4°F depending on the freezer. Even while still sat inside an ice machine, the ice used in most bars is comparatively warm being little under freezing point, 0°C/32°F. Some of the chilling action from the freezer ice will come from its low temperature as heat is conducted from the liquid to the ice, but most of the chilling effect - as much as eight times more - is due to heat energy provided by the liquid being used to melt the ice (known as fusion).
So ice straight from a freezer will chill a drink marginally quicker than ice from an ice-machine, but that difference is small and hardly matters when shaking as thermal equilibrium is quickly reached (as explained above). Consequently ice straight from the freezer has little benefit on temperature when used in a shaker compared to ice from a machine. However, the two different types of ice have dramatically different effects on dilution.
Ice taken straight from a freezer has a dry surface, and if your freezer is cold enough, when touched will stick to your finger as the ice freezes the moisture on your skin. Conversely, even while still sat inside an ice machine the ice used in most bars is wet due to the surfaces of the ice melting - let alone once it has sat in an ice-well for 20 minutes having been carried from the machine in a bucket.
It's obvious that the comparatively dry freezer ice will produce less dilution in a shaken cocktail than the wet ice from a machine. That thin layer of water on each surface of the ice cube immediately adds to the volume of liquid in the shaker, so increasing the ratio of liquid to ice and giving the ice more liquid to chill. Hence the ice melts quicker as it chills the increased volume of liquid producing more liquid. And so it goes on.
So there is a benefit to using ice straight from a freezer and I like to shovel ice from my machine into thick bags which I put in the chest freezer next to the machine. I fill my ice well with ice from the freezer, replenishing the freezer from the machine as I go.
Even if you are stuck with using wet melting ice then you can dry it. Using a salad spinner seems a little over zealous. Instead use a two-piece shaker and pour your ingredients into the small tin (déjà vu from 2 above). Put the ice in the large tin and place a strainer over the ice-filled large tin, upend, agitate and violently lower and raise the strainer covered upended tin to allow most of the water on the ice to drain away. Then introduce the ice to the other ingredients and shake.
So long as your ice is dry (see above) then the size of ice cube has little effect on the level of dilution - be that tubular ice, chunky cubes from a Hoshizaki or Kold-Draft machine, or even boulders hacked from block ice. Crushed ice is another matter...
Shaking with crushed ice will produce dramatically higher levels of dilution than shaking with cubed ice. And the amount of that dilution is hard to predict due to crushed ice having such a large surface area and accelerated melt speed.
A drink shaken with dry surface ice (see above) will gain around 20% dilution from the ice during shaking but some drinks, such as the Daiquiri, benefit from higher levels of dilution. Some bartenders achieve this by shaking such cocktails with a combination of cubes and crushed ice. Levels of dilution produced by this method tend to be inconsistent. It is much better to use ice straight from a freezer and add a measured amount of chilled water as an additional ingredient. Hence, in my recipes it is common to see "½ shot of chilled water (omit if wet ice)" as the last ingredient.
Shaking aerates a cocktail - evident by the tiny air bubbles visible in the drink and the foam that forms on the surface. These air bubbles add a discernible texture not produced by stirring. Sadly these air bubbles dissipate quickly and with them this texture is lost. Hence, shaken drinks should be consumed soon after they have been strained into the glass.
Some ingredients produce and retain aeration, and with it texture, better than others. Try to whip plain water into a foam, or come to that any spirit, and what little aeration is produced immediately dissipates. That's because water molecules are electrically attracted to each other so won't spread out to form bubbles unless something is added that weakens their attraction to each other.
Conversely, whipping or shaking pineapple juice always produces thick lasting foam. Lime and lemon juice also works well but be warned, while straining to remove pulp does not adversely affect citrus fruits' ability to hold aeration, clarification will kill it due to the plants pectin and other polysaccharides being removed.
Milk and cream are also good texturisers, as is milk whey which has the same foaming effect without tasting of milk or curdling when used in acidic drinks. However, due to its high protein content (around 10%) egg white is the king of the cocktail foam.
The size of ice cube you use will have a dramatic effect on the texture of the finished shaken drink - the bigger the cube the better the texture. Shaking with one huge chunk of ice produces the best texture but does not yield enough dilution, so ideally hit a happy medium (as recommended by Dave Arnold) by shaking with one large cube and several smaller cubes. Also consider adding a little egg white to all your shaken drinks.
When making drinks containing cream and eggs it is common practice among many bartenders to first shake the mixture without ice, before shaking the drink a second time with ice. This practice is known as 'dry shaking' and the theory is that first shaking without ice, and so at a higher temperature, better allows the drink to emulsify producing more aeration and a thicker foam on top of the finished cocktail.
Some bartenders also place a spring from a Hawthorne strainer in the shaker during the first 'dry shake' as this acts as a whisk inside the shaker when the drink is shaken.
Dry shaking does indeed produce more foam than conventional shaking with ice. But not as much as 'reverse dry shaking' does. Aristotelis Papadopoulos (Telis to his friends) from Thessaloniki in Greece lays claim to discovering the benefits of this technique. Combine all your ingredients in the shaker and shake conventionally with ice. Then open your shaker and strain liquid back into the smaller tin (supposing you've followed my advice and use a two-piece shaker). Discard ice left in large tin. Reseal shaker and shake again without ice. Then pour your drink into the glass through a fine strainer to catch any curdled egg and the chalaza (the tissue that connects the yolk to the shell's membrane).
Banging a two-piece shaker on the bar to open is decidedly amateurish and when witnessed is usually the prelude to a bad drink. Similarly hitting the shaker with your elbow to break the seal may have once been an accepted flair move but now seems somewhat passé. Even tapping with the palm of your hand is looking slow and dated compared to breaking the seal by pushing the smaller tin diagonally with your thumb. But however you open your shaker, be sure you can confidently open your shaker.