The looksmaxxing ladder rests on one assumption: there is a single scale, and everyone sits somewhere on it. Mog or be mogged. A new dataset from Spencer Greenberg of ClearerThinking.org suggests the ladder is actually several ladders, leaning against different walls.

The quiz

Greenberg posted a “guess the group” game on X: six rows of six men, and the job was to match each row to the group of raters who liked those faces most. On October 20 he posted the answer key. As of this morning it sits at roughly 4.06 million views, 5,042 likes, about 6,230 bookmarks and 767 quote posts. X even turned it into a trending story: “Rugged Man Divides Attraction Preferences in Viral Face Study.”

The faces aren’t real people. All 202 men are AI-generated, and each one was rated by at least 127 people in every group. The raters:

  • Heterosexual women: 65,491
  • Bisexual women: 61,241
  • Gay men: 33,783
  • Bisexual men: 24,415
  • Non-binary people attracted to men: 14,755
  • Asexual people willing to date men: 4,470

That adds up to 204,155 raters. Greenberg’s own read in the replies: “very large sample sizes, so seems quite statistically robust.”

What the rows actually show

This is the part most quote posts skipped. The grid is not each group’s top six. According to the note at the bottom of the image, each row shows “the six men that group rates furthest above the average of the other five groups.” Each rater’s own average is subtracted first, so generous raters don’t skew their group, and every group counts equally, so the 65,000 straight women can’t drown out the 4,470 asexual raters. No face appears twice; one that several groups love goes to the group that loves it most.

In the chart’s own words: “This shows where tastes DIFFER, not each group’s favourites.”

So the rows are a map of disagreement. And there is plenty of it.

Six types

Reading the grid by eye (our description, not ClearerThinking’s labels):

Heterosexual women over-index on the clean-cut, well-groomed guy, often with light stubble or a short beard. Navy T-shirts, neat hair, approachable smile. The faces a looksmaxxing forum would probably rank highest overall.

Gay men go more rugged: broader builds, a couple of fuller-body shots, and the one clear mustache in the whole grid. This is the row the “rugged man” trending headline grabbed.

Bisexual women stand out for longer, darker hair, several shoulder-length, plus beards and more green T-shirts than anyone else.

Bisexual men tilt younger and softer, with a bleached-blond cut and a couple of emo-adjacent fringes.

Non-binary raters attracted to men pick the most alternative row: lilac-grey hair, long curls, a blond mohawk.

Asexual raters willing to date men pick the most varied row, glasses included.

Six groups, one pool of 202 faces, six distinct clusters.

Why it matters for the mog

“Mogging” assumes a referee. Someone walks into the room, and the taller, sharper-jawed man “wins.” But winning requires a judge, and this dataset says the judge’s identity changes the verdict.

To be fair to the ladder, the method is designed to find differences: by construction it surfaces the faces where groups disagree, not the ones they agree on. A face everyone rates highly can still exist; it just ends up in only one row, or none. So this doesn’t prove attractiveness is entirely subjective. It shows that once you control for the shared baseline, the leftovers are big, consistent, and sorted by who’s looking.

That’s enough to dent the forum version of the theory. A canthal-tilt rating out of ten, scored by men on Reddit, is one group’s ladder. The straight women in this study, who are the audience most looksmaxxers say they’re optimising for, have their own standout faces, and they look less like a jawline diagram and more like a guy who owns a decent razor.

Female faces next?

In a follow-up post Greenberg said that if people found it interesting, he’d “do one of these guessing games for female faces, too.” Given the numbers, interest doesn’t look like the problem. Until then, ClearerThinking’s study is still open: take it and you get an analysis of “your type” and your pickiness.

The ladder will survive. It always does. But next time someone posts “he mogs,” the honest reply is now backed by 204,155 people: mogs according to whom?