This article originally appeared on VICE US.
In 2011, OkCupid co-founder Christian Rudder published a set of findings as a now-deleted post on the dating site’s blog. Culled from OkCupid’s own user information, the statistics were cheat codes to figuring out your date in an instant; for example, if someone is tolerant about grammar mistakes, they’re probably at least a little religious—at least, as far as the data goes. The magical questions to ask your potential love match lit up the internet between 2011 and 2012, but it seems that one stat in particular has hung around a bit more stubbornly: Both men and women, gay or straight, are 60 percent more likely to have sex on the first date if they like the taste of beer.
What could liking beer, specifically, have to do with sexual behavior? It’s a question many have asked when learning of the theory, especially women who drink beer.
Although Rudder emphasized this finding was for men, too, it’s been used more frequently to assess a woman’s behavior. Janet Kraayenbrink Viader, who works in sales and operations for a contract brewer in New York, says the stat came up in recent conversation. "[It was] referring to women having this tendency, not men," she said in an email. "It was a man who referenced it. This selective memory definitely plays into our society’s double standard for genders regarding early sexual activity." Not only does this stat tend to be more often repeated by men about women, but it can serve as its own scale of judgment. If a woman reveals she likes beer, will a man she’s on a date with instantly form a certain set of expectations? Will a man try to manipulate his date, purposely pushing for her to order a beer so he can figure out where his night is going?
A classic game of telephone took this piece of data and sculpted it into a gender-biased narrative. An assertion that beer preference is linked to sex on the first date stands out in a study more than an assertion about grammar tolerance and religious beliefs. That stat gets people talking, and they have continued to talk about it for almost nine years—and reference it as fact in an array of articles. This past August, @WorstBeerBlog—a Twitter account known for calling out racism, sexism, and other various levels of stupidity in the beer world—alerted followers to a YourTango listicle, 5 Things You Should Know Before Loving A Lady Who Drinks Beer.
The piece originally included seven things "you should know," and one was that a lady who drinks beer is "DTF." This is an example of how the original piece of data can become skewed. A YourTango editor (who no longer works for the site) assigned the listicle to writer Frank Vamos, Jr. While looking for content involving women, dating and beer, Vamos found the OkCupid stat and included it, but because the article was speaking about women, specifically, he didn't see a need to clarify that the stat was the same for men.
"My initial thoughts were if a popular dating site is making the claim, and I'm writing a piece for a site whose main audience is female, and the tone of the website [...] was light-hearted, tongue-in-cheek and at times goofy, why not throw it in there?," Vamos explained to Munchies.
Vamos wrote the listicle five years ago, but YourTango recycles it sporadically, and it seems that editors first cut the "DTF" entry due to the negative Twitter response after the article’s most recent run in August. You can read the original version here on Thought Catalog. While Vamos was filling his assigned story for—as he described it, a "more clickbait/advertisement-driven" site, some readers might not realize that, and may first learn of this stat in a context like this article that neglects to mention that it pertains to both men and women.
If this alleged link between beer and sexual behavior is going to arise in articles and conversations, it’s worth asking: is it valid? From a scientific standpoint, there’s no reason beer, in particular, would affect someone’s first-date behavior differently than any other kind of alcohol.
"I don’t think people would consider a higher-tannin profile to be a drug—it’s a flavor profile, not a drug profile." said assistant professor at University of Colorado Anschutz Medical Campus Joshua L. Gowin, PhD, comparing this thought process to wine. "The simplest answer is ‘no,’ it doesn’t make sense, because alcohol is the drug and alcohol is the same in every drink."
"All alcoholic drinks are metabolized in the same way, since ethanol is the same molecule in all drinks and is metabolized by alcohol dehydrogenase or the cytochrome P450 enzymes," explained Manfred Brauer, PhD, an assistant professor in the Department of Molecular and Cellular Biology at the University of Guelph in Ontario, Canada. Ruling out behavioral or biological support for this stat, there’s one science left to examine: the data.
Nick Adams, PhD, is a data scientist, sociologist and founder of Goodly Labs. While Christian Rudder could not be reached for comment on his research, Adams was able to make some observations on this OkCupid data. He says there’s reason to believe in its legitimacy, especially thanks to the concept of statistical significance. Often, studies are done on a representative pool—2,000 people to stand in for the entire American population, for instance. Because OkCupid has access to all of its users to conduct studies, its data contains high statistical significance.
However, there are three key things to remember with a stat like this: causation versus correlation, a possible "omitted variable," and something called "ignoring the baseline." First, causation versus correlation: when we read a stat like this, it seems like the link exists in the cause. Maybe something about liking beer does make people more likely to have sex on the first date. But we should be reading a stat like this as a probable example of correlation, not causation. That is to say, among a whole set of factors people who tend to have sex on a first date might have in common, perhaps liking beer just happens to be one of them. It could be a coincidence, or it could have more to do with personality traits and even how people typically respond to questionnaires; this could be the omitted variable.
"There may not be a relationship between whether you’re likely to have sex on the first date and whether you like beer, but in how likely you are to say in a survey that you like the taste of beer, and how likely you are to say in a survey that you would have sex on the first date," Adams said. "You can imagine some other variable like a willingness to share openly."
Finally, we have to be careful not to "ignore the baseline," according to Adams, using theoretical examples for the numbers here.
"If for people who don’t like the taste of beer, maybe 10 percent are likely to have sex on the first date?," Adams explained. "The error that people commonly make is they see the report and say, people who do like the taste of beer are 60 percent more likely [than that 10 percent], so people who like beer are 70 percent more likely to have sex on the first date. You actually have to take 60 percent of 10 percent, which is 16 percent." In other words, these stats shouldn’t be viewed as out of 100, but out of a base rate. The "60 percent more likely" quantifier here refers to 60 percent more likely than the percentage of people who don’t like beer. "Unless we know the percentage of people who don’t like beer who would have sex on the first date, we can’t contextualize that 60 percent," Adams said.
Depending on those numbers, this stat could be quite negligible; liking beer might not actually have a huge correlation with sexual behavior on a first date. Adams said he believes that anyone reporting on a stat like this should be careful to contextualize.
Gowin said he sees how propagating this data without that context could be problematic. "Imagine someone goes, ‘Oh, would like a drink?’ And you say, ‘Sure, I’ll take a wine with my dinner,’ and they say, ‘Are you sure you don’t want a beer?’ to like up the odds," he pointed out. Essentially, this piece of data is most likely rooted in properly collected research, but whenever discussed, it should be partnered with statistical context along with the fact that it applies to men and women.
Really, when it comes to trying to guess if someone is going to have sex on the first date or classify those who might, though, it might be wiser to avoid all "theories."