Monday, September 21, 2026

Game for dating sites

It's amazing how many opinions I have on people forming relationships, for someone who has never dated.

There is the comment I left as a reply to someone on YouTube, 'comment on males becoming mean, 19 Sep 2026.txt', and my reply to their reply. And there's the poll I suggested to Imane in an email about a week ago, which is relevant to the reply I left on YouTube.

I'm not copying those because the purpose of this post is basically, "what if I die of a heart attack tomorrow". Then, even if I had saved the text I just wrote to an (unencrypted) external SSD in case of my hard drive crashing, it would be lost. So that information is already on the Internet; this is not.

Some people might say, we don't need the world to have more people in relationships, because population growth is a problem. I just think of people being unhappy as inefficient. The question from the post Paris, what if all bullets fired in war were lethal? If I could cause that outcome to happen, I would, just to save bullets. (Well, also so people wouldn't have stupid arguments about whether people in war want to kill each other, or whether most of them are firing over the heads of their enemies.)

This is the game:

game for dating sites. Judging someone: doesn't work if there's just one answer that everyone can copy. General solution, making the test hard: education, or 'taking a good photo of yourself'. But depends on good, hard metrics existing. The other general solution: making the test vary. Person 1 (P1): Presents three photos of other people. Person 2 (P2): says whether they think each of these people is more or less attractive than P2. P1: says whether they believe P2.

Iteration: not just a test of appearance. Each photo also has fictional information. P1 can choose from several categories: occupation. Personal details that were beyond their control: "tragic backstory", "has many relatives", "war refugee", "only child", "grew up far from any big city". Degree of romantic experience: high, medium, low. Number of countries they have been to. Up to two categories can be selected for each photo.

If P1's game is to guess whether P2 has lied, then it must be explicitly ok for P2 to lie. P2 selects both true answer, and cover answer. They can differ or be the same. For each guess by P1 of whether P2 was lying, either P1 gets a point or P2 gets a point. Even if P2 lies about whether they lied, P1 can guess this and still get a point.

P1 controls the photos: they can choose photos where the answer is obvious, or try to choose photos where the answer is hard to guess.

Users could choose to allow their own photos to be used for the game by other people. It's made explicit that the populations of male and female are considered to have equal average attractiveness, with a quick graphical introduction to the game. So a player cannot just say, "everyone of the opposite gender is more attractive than me, by definition".


One reason not to post this was additional details that are not relevant to someone who chooses to implement this game.

After I did something I will not provide additional details about at this time, my browser had the display bug. That meant closing my browser, and if I wanted to reopen it, that would be a separate action. Instead, I synced some videos.

I hesitate to say any of this because the government conspiracy is acting a little crazy.

So, point: I'm arrogant enough to say that I think I'm smarter than most people. This is why I left a YouTube comment on people having relationship problems, when I haven't really had the opportunity to ever have relationship problems. (Since my only relationship was with Mei, and we never exchanged photos or talked on the phone, and if Mei saw me in real life while I was living outside in 2012, I'm unaware of it.)

But I would freely say that in the biased sample of videos which I'm exposed to, most people featured in them are more attractive than me. For example, in a video I recently linked, the camera pans away from the stage and the audience is visible. Maybe I could be more attractive than some members of the audience, but I think almost all members of the audience would say that the dancers on the stage were more attractive than the audience member being asked, and I would say this too.

That's also the case in this video, which I felt I had to upload because of the edit at 1:37:

Ttesports Show Girl@2011 TGS台北國際電玩展

Other performances are still on YouTube, like this one:

11.02.20-2011台北國際電玩展Tt舞者開場舞part3.MP4

This one has audio muted on YouTube. Uploaded with original audio and higher video bitrate from when YouTube thought it was a 1440p video, not a 720p video.

Dancer who is on the left and closest to the camera at the start of these three videos is 小姿, full name 黃姿菱 (1, 2). She is definitely more attractive than me. By not mentioning the other dancers I leave open the possibility that I match their level of attractiveness without lying.

In US culture, it's understood that males will often judge the looks of females on a 1-to-10 scale. This is often thought to be a way of insulting females, but here is a comedy skit with a different take.

If this game was to for P1 to ask P2 if the people in the photos were more attractive than P1, it would be bad.

Some people think you should always be honest, even when someone asks "does this clothing make me look fat". Others believe that it's ok to lie to benefit someone else. Anyway, I think this game works. It involves people making judgements based on information they should have good knowledge of: their own appearance. Instead of judgements based on the very basic information they have about someone else's appearance, that on a dating site comes from carefully selected photos etc.


Hopefully that's all that I wanted to say to explain this game before I got distracted and spent nearly two hours looking up videos. (Also, I got the vp9 bug while watching a video.) I was going to say that I thought of it because I was evaluating a video. These people are both more attractive than me:

@gioalbertsec Onda da Balinha (Remix) - Remix - Natralhinha [7136231465607646470][a +0.12s].mp4

@gioalbertoficial Onda da Balinha (Remix) - Remix - Natralhinha [7135822350896942341][a +0.12s].mp4

@diveludo Onda da Balinha (Remix) - Remix - Natralhinha [7136250843199819013][a +0.14s].mp4

But that doesn't make them special, because there are a lot of people who are more attractive than me.

I looked at the first video, and then checked for other videos with the same song because I recognized it. I don't think it's @gioalbertoficial's best videos; I think these ones make her look better: 7153703345285254406 (sync), 7153700589061344518. I actually made a note to myself to check the comments on the first one, if people ever use this idea and I can use the Internet freely without causing unintended consequences.

But no matter how good a real person looks, it's never going to be what a human brain natively thinks of as ideal. (Whereas with repeated exposure, a non-ideal thing can end up being seen as ideal: who thinks about whether the QWERTY keyboard is bad?) In the past, humans could create images that looked better than a real human; now, AI can do so easily.

There are many videos and discussions about how images of attractive people warp our expectations. Even if someone knows that "real people don't look like that", they can't be sure how 'real' a particular photo is. To fall back to the 'ratings system', if someone looks like a 7/10 in a photo, are they a 7/10 in real life or a 4/10? (To put it another way: maybe most people who look like 7/10 are really 4/10, but what if someone is actually 7/10 because they didn't bother to do the things that would make them look 9/10?)

People can end up assuming that other people are lying about or inflating their value, so their interactions — like on a dating site — could be about trying to detect the lie, instead of the conversation that would take place if both parties knew that the other wasn't lying.

Suppose that in the game, P2 wants to convince P1 that P2 is attractive. Do they do so by saying that the people shown in photos are less attractive than P2, or by saying that those people are more attractive?

No comments:

Post a Comment