Skip to main content

Getting Drunk May Be Kind of Like Driving on Cruise Control

Insert morbidly ironic PSA here.

By Nathan Collins


(Photo: Justin Sullivan/Getty Images)

If you’re planning on watching the debate tonight, there’s a decent chance you’re considering drinking your way through it. Of course, the Centers for Disease Control and Prevention generally frowns upon the sort of binge drinking political debates seem to engender these days, so in lieu of whiskey and wine, here’s a good conversation starter: A preliminary study suggests getting drunk may be sort of like driving on cruise control.

It’s a bit of an unfortunate metaphor. As Luis Giraldo and his colleagues write in IEEE Transactions on Cybernetics, heavy drinking kills around 88,000 people in the United States every year, but there aren’t great models to determine how much and how quickly people drink. To fix that problem, Giraldo and his team’s idea was to think of drinking in terms of a proportional-derivative (PD) controller, a feedback system used in an enormous range of electronic control systems including robotics, thermostats, and vehicular cruise control.

It’s not quite as weird as it sounds.

To understand how PD controllers work, imagine you’re driving along and set the cruise control to 65 miles per hour. The simplest cruise control would step on the gas as hard as it could until the car reached 65 mph, but such basic systems have a habit of overcorrecting, so you might actually speed up to 68, then slow down to 62, speed up to 68 again, and so on. A PD system adds two refinements that help prevent overcorrection: The farther you are from 65 mph, the harder the cruise control steps on the gas; and the faster you’re closing in on 65 mph, the more it eases off the gas.

Your likelihood of buying another drink depends not only on how far you are from your drunkenness goals, but also on how quickly reaching those goals.

Now, what could that possibly have to do with drinking booze? Surprisingly, there’s some evidence that how drunk you feel depends not just on your blood alcohol content (BAC), but also on how quickly your BAC is changing, meaning that, after you stop drinking, you’ll feel less drunk than you would if you had the same BAC while still sipping (or chugging) away.

That starts to look kind of like a PD controller: How you feel, and therefore your likelihood of buying another drink, depends not only on how far you are from your drunkenness goals, but also on how quickly you’re reaching those goals.

As a first—and, the authors admit, rough—test, Giraldo and his team compared their model to data from two earlierexperiments that recorded BACs for 1,040 people when entering and exiting a bar, along with how long they were at the bar and how drunk each of them said they wanted to get.

It’s a rough test, since the team didn’t have data on how many drinks participants had or when they drank them, but early indications suggest the model works. For example, entering and exiting BACs were consistent with the idea that people would try to reach their desired drunkenness, although it wasn’t possible to say for sure whether they did so according to the team’s proposed model.

“Currently, we are in the process of collecting real-time drinking data,” to better test and refine their PD-controller model, the researchers write. “Our aspiration is to gain enough understanding to start interventions at both the individual and group event levels.”