As founders of the Seattle-based data science and machine-learning consulting firm Nolis LLC, we analyze information for corporate clients all the time. But we've never had a project like this.

When we got a chance to analyze The Stranger's sex survey data, we were ecstatic. What treasures would we find by using cutting-edge machine-learning techniques on stuff people don't usually talk about? What happens when you merge AI and sexual variability? What kinds of trends emerge?

We live in divisive times, and another thing that clearly divides people is what turns them on. We set out to narrow down people's kinky interests into four basic groups—and to do so, we employed a powerful data-science approach called clustering. A method of unsupervised machine learning, clustering uses an algorithm to reveal new information from the data rather than relying on a human analyst's opinion or agenda. We plotted people in a 32-dimensional space where every dimension represented a different kink, from spanking to blindfolds to bondage to fisting. The k-means clustering algorithm then combed the data and separated people into distinct groups based on how close they were by distance within the 32-dimensional graph.

With clustering, it's essential for an expert to then interpret the groups. The nature of four main groups found by k-means appeared obvious to us, so we named them.

With a powerful method of machine learning called clustering, we narrowed people’s kinky interests into four basic groups.

The largest of the groups we're calling the Squares, because folks in this group prefer vanilla sex—nothing more intense than a playful tush pat (and there's only a 24 percent chance they enjoy that). The Squares constitute slightly more than a third of all respondents (36.3 percent).

The second-largest group we named the Spankers (24.9 percent of respondents), mostly because we wanted to keep up the alliteration (but "the BDSM crowd" works, too). Spankers have a more than 50 percent chance of being interested in spanking, bondage (being tied up or doing the tying), Domination/submission, and blindfolds.

The third-largest group we call the Show-Offs (20.2 percent of respondents). They enjoy voyeurism, exhibitionism, and group sex. It's worth emphasizing that the Spankers (BDSM, etc.) and the Show-Offs (voyeurism, etc.) are distinct groups—meaning few people whipping it out in public also want to whip someone (or be whipped) in the bedroom. Show-Offs are also more into double penetration than Spankers, which makes sense, what with all the group sex Show-Offs are having.

The fourth and dirtiest group of respondents (18.5 percent) we're calling the Kitchen Sinks, because they're into everything. They want the works. Not only are they heavily into BDSM, voyeurism, Dom/sub, double penetration, and blindfolds, but they also have a more than 50 percent chance of being into pain and rape/intruder role-play, and a nearly 50 percent chance of being into asphyxiation and fisting.

If our kinks divide us, maybe sexting is what connects us.

We wanted to understand the relationship between what dirty pics people send and what dirty pics people want to receive, so we constructed Sankey diagrams that map, on the left, what people send, and on the right, what they want to receive. Turns out, people want straight-up pussy and dick pics a lot more than they are willing to send them. People who send butt pics are unlikely to want butt pics, and vice versa. People who send boobs want dick the majority of the time, and are more likely to want boobs back than pussy. People who send dick pics want a healthy split of other parts, but always want something. Most interestingly, this analysis identified a curious, greedy sect of people who refuse to send dirty pics themselves ("I don't send dirty pics, sluts!" was the option they selected on the survey), but who absolutely expect you to send your nudes.

People who send boob pics mostly want dick pics. People who don’t send pics want to see boobs more than other body parts.

Those selfish sexters comprise a whopping 15 percent of responders. Who are they? Would you be shocked to learn that selfish sexters are vastly cisgender men? According to the numbers, when dealing with cis men, there's a 30 percent chance the dude won't cough up his own nudes. At the other end of the spectrum, cisgender women have a less than 5 percent chance of holding tightly to their self-smut. Trans men behave closely to cisgender men, and trans women closely to cisgender women.

Finally, we wanted to see how sex habits and experiences differ by age.

We used a popular quantitative research technique called crosstab analysis, which means that we split the data by one variable and look at another. Used responsibly, this is the method by which researchers can determine the answer to questions like "How does your location influence how you vote?" Used irresponsibly, this is the method by which nosy people like us can determine if your chance of being into group sex changes with age. It doesn't. Young and old people have roughly the same chances of being into group sex.

Likewise, young and old people had the same chances of having used flavored lube or having had their butt fucked. But that's where the similarities stop. When it comes to having sex while watching Game of Thrones, people under the age of 36 were four times more likely to have done that than people over the age of 55, likely due to how much "the youth" are into the show. People under the age of 36 were 2.8 times more likely to be into findom (financial domination) and 3.2 times more likely to be into asphyxiation than people over 55. Meanwhile, people over 55 were 1.9 times more likely to be into amputee sex and 1.6 times more likely to be into feet than their younger counterparts. This suggests that kinks go in and out of style just like other parts of culture.

When it comes to drugs, people under the age of 36 are 1.3 times more likely to have had sex on ecstasy/Molly/MDMA than people over the age of 55. By contrast, people over the age of 55 are twice as likely to have had sex on poppers, 2.4 times more likely to have had sex on meth, and 6.1 times more likely to have had sex on Viagra.

And what about having sex at Golden Gardens, the scenic beach in Ballard? What about having sex on private boats? People under the age of 36 were 1.2 times more likely to have had sex at Golden Gardens, but people over the age of 55 were 1.9 times more likely to have had sex on a private boat.

We also looked at how age factors into two of the more shocking questions on the survey: the question about peeing on Donald Trump and the question about which animal respondents would sleep with if they had to sleep with one.

In general, people under the age of 36 want to fuck a centaur a majority of the time and pee in Donald Trump's mouth—but once you go over the hill, the chance of that being true drops dramatically. Survey respondents over the age of 55 were more likely to say they wanted to pee in Donald Trump's hair and fuck animals besides centaurs (dolphins and dogs ranked highly). Older people were also more likely to skip the question entirely than younger people were.

Despite everything we've just told you, we've barely scratched the surface of the dark, deep intricacies of your sex lives. We were intellectually invigorated by your sexual vigor—and we hope you are, too. Thank you for your honesty, your time, and your trust. If you have any other data you'd like us to assess, our services are available. Get in touch through our website at