PicoBlog

How hot is a 5, actually?

We have the classic 4chan image:

But is it true? Is this an accurate assessment of hotness?

I decided to run a survey to find out.

I collected ~240 images of women’s faces, most of them AI generated. I just default scraped a bunch of them, with the only stipulation being that they had to look reasonably above 18, and had to be plausibly fertile (with only a few exceptions).

Some of the non-AI generated ones I took from a public directory of images; a few were submitted by some of my female friends.

Survey was simple - you get an image, and rate it on a scale from 1-10 on how attractive you think the woman is.

As I write this, an average of ~1200 men have rated each face. More are pouring in cause this was a real popular survey holy shit, but 1200 seems really plenty to get high confidence. I’m only looking at ratings from men, for this post.

I normalized the results, because you can’t really get an average of 1 if the lowest option possible is 1. This had the effect of making the hottest woman’s score a 10, the 2nd hottest a 9.2, the 3rd hottest a 8.6. So I’m just rounding the “top hottest” into the ranking of 9, which is the highest non-personal score possible on the 4chan chart.

These are also ones below a certain standard deviation threshold - which means if people really disagreed in their ratings, the photo is less likely to be included here.

So here’s the official How Attractiveness Ratings for women:

In keeping with 4chan tradition, there are no automatic 10’s; someone from the 9 category might become a 10 for you, if they meet a specific niche you like a whole lot.

The above are adjusted ratings, so that the lowest-rated is automatically a 1, and the highest rated automatically a 10, and the other photos distributed proportionately between.

But raw ratings tell a different story.

On average, men rated this photo at 8.4/10 attractiveness, which was the highest score given to any photo in my collection.

I googled ‘world’s most beautiful woman’ and the internet told me Yael Shelbia has won some awards and made a bunch of lists. To help calibrate, I included two images of her in the dataset.

Both got almost exactly the same score, which is pretty cool - 7.3. (Her normalized score was 8.5)

I often see the okcupid graph showing men rate women as a pretty even distribution of attractiveness, but my graph shows that distribution skewed left. Here’s the histogram of average ratings assigned each photo:

And for fun, here’s the most exactly 5/10 rating photo, for both normalized and not ratings:

A few more things to go over:

But first, here’s the ones women thought were hot, and men did not:

ncG1vNJzZmiZlaG5onrSrpmsrJGYuG%2BvzqZmqWenpLqmutJmn6isnprAtHnSnJilnQ%3D%3D