Some bartenders have a chip on their shoulders when it comes cocktails. Shaking waters it down, they say. Stirring is better. James Bond is a pussy.

We wonderedâ€”what's the scientific difference? Does one way or the other produce a pour that's more warm or watery? Do the fluid dynamics in the darkness of a shaker lead to other intangibles that make our favorite drinks so delicious? Prepare to have your faith shaken, because we have reached some stirring conclusions.

It's Friday afternoon, you've made it through the long week, and it's time for Happy Hour. Gizmodo's weekly cocktail column is a bottomless bottle of innovation, science, and alcohol. "No, Mr. Bond, I expect you to drink."

The Question

What inebriates better: a cocktail that is shaken or stirred?

The Methodology

We've seen subjective tests of the shaken vs. stirred debate beforeâ€”the Mythbusters confirmed that you can taste the difference. But is that due to temperature, alcohol by volume (ABV), or something else? What about all those ice chips you see in a shaken drinkâ€”do they matter?

We began the test by measuring all the ingredients by weight, starting with the ice. Different sized ice cubes have larger or smaller surface areas, and thus melt differently. So we made 14 identical ice cubes in the same tray, each using exactly 25 grams of water.

For the alcohol, I measured out 70 grams of terrible, cheap, 80-proof vodka (40 percent ABV). A digital thermometer indicated the vodka's temperature to be 73 degrees Fahrenheit.

To calculate the contents of the shaken and stirred samples, we used a digital scale to measure how much water weight the solution gained, and a distilling hydrometerâ€”called a Proof and Tralle Hydrometerâ€”to measure proof. This thing is pretty cool: It floats in a vial of liquid, with preset levels corresponding to the solution's ABV. We tested the hydrometer's accuracy by checking still water, then uncut Georgi vodka, and it dialed up dead-on measurements of 0 proof and 80 proof, respectively. Excellent.

The Experiment

To create solution A, our shaken sample, we poured 70 grams of vodka (just under 2.5 fluid ounces) into a standard metal cocktail shaker. We added five 25 gram ice cubes all at once, slammed on the lid and shook the hell out of it for a stopwatch-timed 30 seconds. We immediately strained the drink into a dry, room temperature glass and inserted a digital thermometer. We set the drink aside, covered it, and allowed it to get back to room temperature for an accurate hydrometer reading. Then, to the scale for a weight check.

To make the stirred solution B, we repeated the previous stepsâ€”70 grams of vodka, five 25 gram ice cubesâ€”and then gently but thoroughly stirred with a long spoon for 30 seconds. The drink was strained into a separate, identical glass, then checked for temperature, proof, and weight.

The Results

â€˘ Temperature: This was the first dramatic difference. The shaken drink dropped down to a frosty 29 degrees F, whereas the stirred cocktail measured only 38.1 degrees. The ice cubes couldn't have transferred all that heat without melting and therefore diluting the solution. Time for a weigh-in.

â€˘ Weight: Both solutions started at exactly 70 grams of pure vodka. After stirring the liquor with the ice cubes, solution B gained 16 grams of water weight, coming in at 86 grams. That sounds significant, until it is compared to the weight of shaken solution A: 116 grams! It gained a whopping 46 gramsâ€”more than half its original weightâ€”from trace amounts water knocked off of the ice cubes.

â€˘ Proof: Now, for the hydrometerâ€”the moment of truth. Once both solutions had reached a temperature of exactly 72.4 degrees F, I tested them, twirling the hydrometer to free any lingering bubbles. The stirred drink had dropped down from 80 proof to a count of 60, or 30 percent ABV. Now, listen to this: The shaken drink's purity had plummeted, with the hydrometer hovering between the lines marking 45 and 46 proofâ€”around 23-percent ABV. In other words, shaking just ice and alcohol can cut a spirit's potency nearly in half, and which dilutes a drink 1.75 times more than stirring it does.

We cross-checked this with the amount of water weight gained, and it adds up. According to the weight, the stirred Solution A is now 32.6-percent ABV (65.2 proof) and the the shaken Solution B is now 24.1-percent ABV (48.2 proof). In other words we're within a two percent margin of error, which ain't bad at all.

I passed the glasses around the office for a subjective taste test, and everyone agreed that the stirred drink tasted much stronger. And this crowd knows what a cup of lukewarm Georgi is supposed to taste like. So there you have it.

So should I shake or should I stir?

Economical drinkers usually want the most booze for the buck, so it would seem stirring is the way to go. Savvy readers may remember that last week's Happy Hour on booze myths stated that diluted drinks actually get you drunk faster than straight drinks. But a martini glass is only so big. In bars, what doesn't fit into a glass gets poured out. A 4-ounce martini that's 30 percent water simply has less vodka in it than a 4-ounce martini that's 15 percent water. So forget James Bond and his "shaken, not stirred" mantra. He was probably just pacing himself so he could fight/screw someone in the next scene.