Big data and big dating: The algorithm of attraction
Perspectives > Blog Post

Big data and big dating: The algorithm of attraction

March 15, 2015

We all know someone who has met their significant other online, or have heard some of the tragic/romantic/off-the-wall stories of someone who has dated online. At the very least, you’ve seen the proliferation of online dating sites such as OKCupid, eHarmony, Tinder, Hinge, and the more tailored platforms such as JDate, ChristianMingle, and FarmersOnly.com (yes, this exists). A recent Kelton study about online dating found that a third of Americans have used on online dating site or app – that’s three times as many people when compared to stats from 2013. Yes, that means over 80 million people have likely poured hours into crafting their profiles while questioning every detail – is this funny? Do I come off as well-traveled? Is it ‘your’ or ‘you’re’? How many selfies are too many? And it’s not just those digitally-savvy Millennials; about 1 in 5 adults age 45+ have used digital dating sources to find love.

We all know someone who has met their significant other online, or have heard some of the tragic/romantic/off-the-wall stories of someone who has dated online. At the very least, you’ve seen the proliferation of online dating sites such as OKCupid, eHarmony, Tinder, Hinge, and the more tailored platforms such as JDate, ChristianMingle, and FarmersOnly.com (yes, this exists).

A recent Kelton study about online dating found that a third of Americans have used on online dating site or app – that’s three times as many people when compared to stats from 2013.  Yes, that means over 80 million people have likely poured hours into crafting their profiles while questioning every detail – is this funny? Do I come off as well-traveled? Is it ‘your’ or ‘you’re’? How many selfies are too many? And it’s not just those digitally-savvy Millennials; about 1 in 5 adults age 45+ have used digital dating sources to find love.

1

With millions of love-seeking singles pouring their hearts (and smartphone batteries) out, big data has never been so fun. Christian Rudder, co-founder of OKCupid, would agree, giving evidence within his book, Dataclysm: Who We Are When We Think No One’s Looking. This mathematically-based narrative of virtual loveoffers an intriguing and calculated look at the behavioral data of American singles on OKCupid, one of the largest online dating platforms in the US.

One hypothesis going into his data analysis was that the rise of things like airbrushing, cosmetic surgery, Bud Light ads, and prominence of women like Kim Kardashian in the media has set unrealistic expectations when it comes to what men think is attractive. His findings show that not only are men’s expectations and attractiveness ratings of women ‘normal’ (literally and statistically), it’s actually the women who have higher or more complex expectations of men. The distribution of women’s attractiveness ratings of men are skewed much lower, meaning they are much less likely to rate men with high attractiveness scores (4 or 5) and on average receive a rating of about 1.8-1.9 (compared to men who rate women with an average of about 3.1-3.2).

Findings from Dataclysm:

2

Copyright 2014 Christian Rudder. Used under Fair Use.

This is fascinating, and naturally leads one to ask ‘why’, along with several other questions and hypotheses about how men and women calculate attractiveness –Are the men on OKCupid just really unattractive? Is it just about appearance for men? Are women taking more factors into account? Do they have more deal-breakers? Unfortunately, big data can sometimes leave us with a surface-level view of behaviors and doesn’t enable us to understand the ‘whys’ behind them. Christian conducted tests to show that the answer to the first question is ‘no’ – the men on OKCupid are not abnormally unattractive. And a recent Kelton study about online dating helps us shed a little bit more light on some of the other questions (and a few others!):

1.     Men are not just ‘all about looks’.

For both men and women, appearance was the number one factor in determining whether a person is “attractive” – this is not surprising. But this is not the only driver – for men or women. On average, men take into account about 6+ factors when determining the attractiveness level of someone, including non-physical features such as age (43%) personality (39%), the type of relationship they are looking for (37%).

2.     Spoiler Alert: Women are complicated.

On average, women use a more complex algorithm, taking into account over 10 factors when determining how attractive someone is (compared to the previous 6 for men).

3.     Correct grammar usage is sexy.

Hitting the books is more important than the gym. For women, grammar usage (48%) is a bigger driver of attractiveness than body weight (42%) or height (40%) on average.

4.     Quit it with the selfies.

For more than half of women (54%), the type of photos posted by others (e.g., number of selfies, whether they seem exciting/funny/adventurous, the type of clothing they are wearing) is a key driver in determining how attractive they are – again, even more important than weight and height.

So it turns out that women are a bit more complex – at least in their attractiveness calculations. And finessing (and proofreading) your online dating profiles is worth the time investment for everyone. Go get ‘em, tiger.

 

 

 

 

 

 

 

 

Up Next