Much Ado About Stirring Time

I tossed a lot of vodka down the drain last night. But not for the reasons you may first think. There was a question that was keeping me awake at night. A question with ominous overtones. A question that, to my chagrin, lacked an answer having real consensus.

How long should I stir a cocktail?

An oft overlooked, yet essential, ingredient to any cocktail is ice. Ice adds chill. Ice adds dilution. One cannot be had without the other. A wide range of recommended stirring times are cited across cocktail books, forums, blogs, and e-zines in order that optimal temperature and dilution is achieved. Jason Kosmas and Duschan Zaris suggest 40 revolutions. Based on my “stirring speed”, if you will, this amounts to about 18 seconds of stirring. Jim Meehan, in his masterful PDT Cocktail Book, states that his crew stirs gently for 10 to 13 seconds. Others say 30 seconds. Forty seconds. A full minute! In short, recommended stirring times are all over the shop.

Maybe I should take a step back. What exactly is “correct” or “optimal” temperature and dilution? I think we can fundamentally agree that a cocktail needs to be chilled and diluted to some degree. How cold? How much dilution? I suppose this largely comes down to personal preference. Some like it boozier than others (I think I might fall firmly into this camp). There’s a rule of thumb being bandied about that suggests that a well made cocktail should contain around 25% water. Based on absolutely no empirical evidence whatsoever, this number seemed a tad high to me, but I decided to proceed on this basis as a helpful benchmark.


It’s evident that stirring time is not the sole influencing factor effecting temperature and dilution – stirring speed, type of ice, and the initial temperatures of the mixing glass, ingredients and ice are also expected to contribute.

With regards to stirring speed I counted the number of revolutions while stirring for 60 seconds and repeated 3 times. My average stirring speed is 135 rev/min, or 2.25 rev/s (quicker than I would have guessed).

As for ice, I use Tovolo’s Perfect Ice Cube Tray with filtered water, resulting in cubes that are approximately 32 x 32 x 32 mm (1¼ x 1¼ x 1¼”) in size. For this rudimentary experiment, I let the water freeze for a full 24 hours prior to mixing. I typically use 6 cubes per cocktail, which fills my mixing glass ⅔ to ¾  full. Six cubes of ice amounts to 192 grams (about 6¾ ounces) on average.

For this modest experiment, I chose to mix vodka martinis. Why vodka martinis?

    1. Vodka and dry vermouth can be had for pretty cheap; and
    2. I see no point in vodka, I dislike vodka, and, therefore, don’t feel terribly upset about tossing it out (in the name of Science, of course).

I used the following recipe:

    • 2¼ oz. 80 proof vodka (room temperature, approx. 20°C)
    • ½ oz. dry vermouth (refrigerated, 4.7°C)

Bitters and garniture were considered superfluous for the task at hand and omitted for the sake of simplicity.

The rest of the apparatus employed in this cocktail science endeavor was as follows:


The method used to determine the effects of stirring time, whilst holding all other factors reasonably constant, was as follows:

    1. Place mixing glass oh-so-gently onto digital scale and zero;
    2. Add vodka and vermouth to mixing glass and record mass of ingredients prior to stirring;
    3. Measure and record the initial temperature of the vodka and vermouth mixture, taking care not to touch the interior surfaces of the mixing glass;
    4. Take 6 cubes of ice straight from the freezer and add to mixing glass;
    5. Insert spoon into mixing glass and initiate stop watch;
    6. Stir for a pre-determined time;
    7. Remove Baby Bullet Short Cup from freezer, place on scale, insert thermometer and zero;
    8. Strain cocktail into Short Cup;
    9. Measure and record mass and temperature of cocktail;
    10. Empty stirring glass and allow to return to room temperature;
    11. Discard vile vodka martini, rinse cup and return to freezer;
    12. Repeat for other stirring times.

Figure 1 shows the general experimental apparatae.

Paternal Drunk - Post 14 - Stirring Scale

Figure 1: Stirring Time Experimental Apparatus

The above described procedure was repeated for stirring times of 20 seconds, 30 seconds, and 60 seconds. At the end of the battery of tests, I transferred the last cocktail to a proper glass, and added a couple dashes of orange bitters and a lemon twist, but alas, these additions could not ameliorate this drab, abominable concoction. Tasting notes: citrusy nose giving way to the faintest hint of apple or peach and then nothingness. Bland and forgettable. Down the sink. Our taste buds deserve better.

Table 1 indicates my complete findings, showing initial mass, mass at the end of stirring, change in mass (i.e. mass of water added), initial temperature, temperature at end of stirring, change in temperature, and volume of water added (calculated using water density corrected for temperature).

Table 1: Stirring time results
Trial Time
(s)
Mass (g) Temperature (°C) Volume of Water Added Percent Water
(%)
M0 M(t) ΔM T0 T(t) ΔT (mL) (fl oz)
1 20 73 91 18 19.4 1.2 -18.2 18.00 0.61 19.8%
2 30 72 93 21 18.8 -0.3 -19.1 21.00 0.71 22.6%
3 60 70 99 29 18.3 -3.8 -22.1 29.02 0.98 29.3%

I typically stir for 20 seconds. Maybe 25 if the spirits in use are of a particularly hair-growing high proof. Based on the data in Table 1, this results in adding 0.61 to 0.66 oz. of water to the cocktail (19.8% to 21.2% H2O/vol., respectively) and reducing the temperature to between 1.2 and 0.5°C. This seems to suit my tastes just fine. This being said, if I were to achieve the 25% H2O/vol. baseline, I would need to stir for about 42 seconds (looks like a linear trendline fits the data fairly well – see Figure 2). One thing is for certain: stirring for a full minute yields a cocktail, which – to my taste, anyway – is too watery (albeit mighty cold). For those of you who prefer to absorb their information visually, I give you Figure 2.

Paternal Drunk - Post 14 - Stirring Graph

Figure 2: Effects of Stirring Time on Dilution and Chilling

So, what gives? Do the famed bartenders mentioned in the intro not know what they’re talking about when they suggest stirring times ranging from 15 to 20 seconds? That seems highly doubtful, indeed. In the YouTube videos I’ve seen, the size of ice and stirring speeds they use seem to be inline with what I have used in my testing. So perhaps the 25% H2O/vol. guideline is bogus? That’s what I am leaning towards; I don’t think there’s a single ideal dilution that can apply to every stirred drink. Again, it may be that it boils down to personal taste. That being said, I think I’m going to start stirring my cocktails a tad longer moving forward. Thirty seconds seems like a good balance between chill and dilution. Thanks, Science!

I welcome your comments on this topic of cocktail-geekery. Word to your Mother.

Advertisements