This is the probably the most sophisticated, data-driven, scientifically backed gender attractiveness inequality visualization you’ll ever see:
To fully understand it, we first need to see how it was made.
The more detailed US dataset contains teachers’ ratings of their students’ attractiveness (on a 5-category scale) at age 7, 11, and 16. Here is the US dataset extrapolated and visualized as a cumulative graph.
As you can see, girls are more attractive at essentially every percentile. Attractiveness ratings overall are quite generous, as only about 20% of children are rated “unattractive” (visualized on the y-axis between 2 and 4) or “very unattractive” (visualized on the y-axis between 0 and 2).
Now, I don’t find the aforementioned datasets particularly interesting or useful for two reasons. Firstly, I am primarily interested in beauty as it relates to heterosexual relationships, and I don’t find childhood attractiveness applicable (or even appropriate) to study in this context. Secondly, these datasets contain intragender ratings (males rating males; females rating females), which aren’t actually relevant in heterosexual relationships, where the only thing that matters is how men rate women, and how women rate men.
All platforms find the same result: women rate male attractiveness lower than males rate female attractiveness
The following research graph visualizes the data from the “Unspecified dating platform” (the researchers of the study did not disclose which platform they used to gather data).
OkCupid, Unspecified Dating Platform and HotOrNot agree on the basic finding that women rate male attractiveness lower than males rate female attractiveness, but OkCupid provided the most detailed information, including 7 data points for both sexes, visualized in a normal distribution graph. I turned the normal distribution to a cumulative graph, which in mathematical terms is called a Lorenz curve.
The original ratings were provided on a 7-point attractiveness scale, which I scaled and extrapolated to an 11-point attractiveness scale, from 0 (least attractive) to 10 (most attractive), such that 5 is the median.
This is how it turned out.
The data reveals a few interesting things.
Firstly, female attractiveness (as rated by men) is very fair. Most women are around average, with few being very ugly and few very beautiful. Furthermore, women are equally likely to be above average looking as below average looking. The mathematical mean and median are basically exactly at the 50th percentile.
Male attractiveness (as rated by women) on the other hand, is very unfair. Women rate an incredible 80% of men as below average looking.
In effect this means a 50th percentile woman is as beautiful as an 80th percentile man.
Now this raises a question. If a 50th percentile woman is considered as beautiful as an 80th percentile man, does this mean the median woman will actually pursue an 80th percentile man?
To answer at this, we can look at data on aspirational pursuit. Aspirational pursuit is basically the tendency to seek out partners more desirable than oneself. One study found that people aspirationally pursue about 21%  [note:3], however, this is the average of both genders. Thankfully there are two better datasets on this phenomenon, and thankfully they both define desirability in percentiles as opposed to rated-attractiveness.
Here’s how far up in desirability-ranking people pursue:
If we combine the data and average up, men aspirationally pursue 22% and women 17%. Of course, these numbers are averages. An 80th percentile man cannot aspirationally pursue 22 percentile points up.
Women pursue men they consider worse looking than themselves. This means women don’t necessarily pursue their so-called “looks match”. This is in line with data from old-school dating website hotornot.com, where researchers found that “female members were significantly less influenced by the consensus physical attractiveness of their potential dates than male members were” . Meanwhile, the story for men is completely different.
Men pursue women significantly more beautiful than themselves. Perhaps this seems obvious, Given the widely reported finding that males focus more on physical attractiveness in mate selection than females do. Aslo, men are much less influenced by their own desirability. For instance, one study found that “men’s self-worth was not related to the popularity of the women they contacted” 
Aspirational pursuit has a profound effect on the dating world. Unattractive people are being aspirationally pursued away from, whereas attractive people are being aspirationally pursued to. This is the equivalent of “the rich get richer, the poor get poorer”.
But how unequal is the online dating market exactly? Here’s the list of datasets I used to visualize this. Some have already been mentioned, some I will explain more.
|Okcupid ||7-category scale||7 data points|
|Tinder (swipestats) ||Percentile||3 data points|
|Tinder (blogger) , ||N/A||Gini coefficient (male only)|
|Hinge  [Note:4]||N/A||Gini coefficient|
|Unspecified online dating platform *||7-category scale||Mean and standard deviation|
|Childhood US [Add Health]*||5-category scale||5 data points|
|Childhood UK [NCDS]*||2-category scale||2 data points|
We can use something called gini-coefficients to quantify inequality. Gini coefficient is a number between 0 and 1, where 0 corresponds with perfect equality and 1 corresponds with perfect inequality. For reference, the gini index of US income is about 0.42; for income in Scandinavia it is about 0.28. Basically, income is more equally distributed in Scandinavia than in the US.
Analysts from the dating app Hinge calculated the gini coefficients to 0.542 for males, and 0.324 for females. The original article has since been deleted, but this is what that looks like.
A blogger also calculated the gini coefficients male face on Tinder to be 0.58. This is even more unequal than Hinge. The Tinder economy has more inequality than 95.1% of all the world’s national economies.
Unfortunately, the aforementioned blogger didn’t calculate the Gini coefficient of female Tinder users. Here’s the Hinge and Tinder “dating economies” overlaid eachother.
Now, I wasn’t quite satisfied comparing the female-hinge graph to male-tinder graph. I collected even more data (male and female) from swipestats.io. The male side of the swipestats data corroborates the findings from the bloggers experiment, which is good. The data is significantly more accurate than the blogger, but also much more limited. This means the following graph is heavily extrapolated, and should be taken with a grain of salt. That being said, I still view it as quite accurate in light of all of the other findings previously discussed in this article.
Finally, this is what it looks like when we combine all of the previous graphs. That includes data from: OkCupid (rated attractiveness), Hinge (gini coefficients), Tinder (male gini coefficient), and Tinder (swipestats).
This brings us back full circle. By filling the upper- and lower standardized bound by gender, and adding an “average” line to go with it, we get the original graph that I presented at the beginning. I genuinely believe this to be the most accurate, data-driven, objective, balanced view of the gender attractiveness inequality.
Now, you can reasonably make the argument that online dating paints a distorted view of reality. The problem however, is that online dating makes up a significant portion of reality. In fact, about 40% of heterosexual couples met online  in 2020.
It is true that choice overload causes women to become “increasingly likely to reject potential partners, while for men this effect was either weaker, similar, or not significant.” . That being said, women are still more selective than men even when conditions don’t favor them. One study looked at an Israeli dating service wherein women outnumbered men 1.55 to 1, which found that women are still more selective than men .
Just for fun, this is what the combined graph would look like if we include UK childhood ratings, as well as the highly extrapolated Unspecified Dating Platform study.
These researchers did not disclose which dating platform they used for their researchers. They state that women’s photos were rated to be more attractive than men’s on average. However, this may not exclusively be heterosexual ratings as the paper states ratings came from “Judges recruited from the authors’ departmental email lists rated the attractiveness of the photos belonging to the gender they preferred to date using a 7-point scale (−3 = very unattractive, 3 = very attractive).” so there may have been some homosexual ratings. The visualization makes a lot of assumptions, and is highly extrapolated. This is the only data that I used to make the graph: “In the current sample, women’s photos were rated to be more attractive than men’s on average: Mean of women = −0.02, SD = 1.06; Mean of men = −0.61, SD = 1.05, t(964) = 8.54, p < .001”. That is why it is not included in the final graph.
Original data came directly from OkCupid themselves, but got deleted after significant controversy. Linked is an archived version. The Original link was: http://blog.okcupid.com/index.php/2009/11/17/your-looks-and-online-dating/
21% was calculated from Initiators’ mean attractiveness versus their contacts’ mean attractiveness. Attractiveness ratings came from independent judges, and had intergender standardization.
Original data from Hinge themselves, but have been deleted after controversy. Googling “hinge 15% of men receive 50% of likes” yields tons of results about it. Original link address was: http://hingeirl.com/hinge-reports/whats-the-biggest-challenge-men-face-on-dating-apps-a-qa-with-aviv-goldgeier-junior-growth-engineer/