No products in the cart.
APPLIED FIZZICS · Field Notes · Edition No. 14
A whimsical look at the universe through a glass of Champagne (or your favorite cocktail).
October 27, 2026 · Seattle, Washington
Dear Reader,
James Bond famously preferred his martinis shaken, not stirred. He also preferred vodka to gin—an offense so egregious that, for the purposes of this experiment, we're simply going to pretend it never happened.
But shaken versus stirred? Now THAT is a question worthy of scientific investigation.
Does shaking make a colder martini? Does it add more water? Or is the whole thing just a matter of personal preference?
Well, there's only one way to find out.
<
The Experiment
We prepared two identical martinis, each containing 80 mL gin and 20 mL dry vermouth.
Each received 100 grams of ice. We stirred one and shook the other, using our newly built Raspberry Pi temperature logger to record what happened.
After mixing, we strained each cocktail into a graduated cylinder to measure its final volume.
The Results
| Measurement | Stirred | Shaken |
|---|---|---|
| Initial Volume | 100 mL | 100 mL |
| Final Volume | 135 mL | 145 mL |
| Dilution | 35% | 45% |
| Lowest Temperature | −3.25°C | −3.50°C |
We define dilution here as the increase in volume divided by the original cocktail volume. So, for example, if the original ingredients totaled 100 mL, and the final cocktail was 130 mL, we call that a dilution rate of 30%.
Shaking produced 10 percentage points more dilution than stirring.
But the temperature measurements revealed something equally interesting.
The shaken martini reached 0°C in approximately 15 seconds, while the stirred martini took about 26 seconds, measured from the beginning of their recorded cooling curves.
Eventually, both cocktails approached the same temperature, roughly −3°C.
Shaking cooled the martini faster. Stirring got it nearly as cold, but took longer.
Physics Corner
Where Does the Water Come From?
It is important to understand that ice does not transfer "cold" to the beverage. Rather, the ice absorbs heat from the beverage, reducing the average thermal motion of the molecules in the liquid.
Some of that energy warms the ice to its melting point, and additional energy melts it.
The melted ice becomes part of the drink.
That's why dilution isn't simply an unfortunate side effect of making a cold cocktail. It's an inevitable consequence of chilling a drink with melting ice.
Shaking agitates the ice and liquid vigorously, increasing the rate of heat transfer. Stirring achieves the same basic process more gently.
In our experiment, shaking both cooled the cocktail faster and produced greater dilution.
So Was James Bond Wrong?
Not necessarily!
If your objective is to chill a martini rapidly, shaking certainly has its advantages.
But if you prefer a clearer, silkier cocktail with less dilution, stirring may be your method of choice.
Ultimately, the best martini is the one you enjoy drinking.
Even if you insist on making it with vodka.
We have a new quiz for you this week:
Until next time, stay curious.
Cheers,
Evan Wallace
President, Applied Fizzics Inc.
Makers of The Perlage System®
Want more Champagne science, carbonation experiments, and assorted investigations? Visit the Applied Fizzics Field Notes archive .
Ok, this one leaves a question. If you were to shake and stir down to zero, and measure immediately at the zero mark instead of the set 30 seconds time, would the dilution rate change? The shaken one would be 10 seconds shorter, which I think would make for less dilution than if measured after 30 seconds. If shaken is faster and has about the same or less dilution because the ice was mixed with the liquid for less time, it would be better and Bond would be right. Also, if I were at a bar and they were busy, I wouldn’t expect my martini to be cold after likely an equal inadequate time of shaking and/or stirring.
Daniel Salenski:
Excellent question! And I think you’ve identified an important distinction between two different experiments, one of which we did and one of which we didn’t.
We compared shaking and stirring for approximately equal periods of time. You’re suggesting that we instead mix each cocktail until it reaches the same temperature, regardless of how long that takes.
That could produce a very different result.
Shaking transfers heat more rapidly, but the total amount of heat that needs to be removed to reach a given temperature is approximately the same either way. Since melting ice absorbs much of that heat, we might find that dilution is also similar when both drinks reach the same temperature.
In other words, shaking might get us to the desired temperature much faster without necessarily producing more dilution!
I think you’ve suggested a very worthwhile follow-up experiment. We now have the instrumentation to test it.
And you’re quite right about a busy bartender. Ten seconds of stirring and ten seconds of shaking aren’t necessarily equivalent preparations.