Why the traditional face rating system is flawed

[Sadist]

For me it still feels pretty arbitrary and not even very accurate. And the division into those categories also seems a little out of the blue und unfinished

its 40h 25a 20d and 15m on ca, but there are even more advanced systems nowadays, where the weights have been further re-evaluated and are distributed differently, with other cool caveats that further improve the objectivity of it, but which i am afraid i can not disclose (at least for now)

 

[Tensor]

its 40h 25a 20d and 15m on ca

ok. I'd weigh them quite differently tho tbh

, but there are even more advanced systems nowadays, where the weights have been further re-evaluated and are distributed differently, with other cool caveats that further improve the objectivity of it, but which i am afraid i can not disclose (at least for now)

cool, I'd love to know if you if/once you are willing to share. I could imagine alot more nuance than the traditional system and am happy to hear that it is getting innovated!
Also idk if this just means the weighting is adjusted but I'd rework/remove categorisation into the 4 categories that are individually weighted and add some things i feel like are quite disregarded, like shapes and countours.
Also I'm beginning to think the result of a rating should not be a set score but a distribution a a scale

 

[Sadist]

cool, I'd love to know if you if/once you are willing to share. I could imagine alot more nuance than the traditional system and am happy to hear that it is getting innovated!

will do if i remember

Also idk if this just means the weighting is adjusted but I'd rework/remove categorisation into the 4 categories that are individually weighted and add some things i feel like are quite disregarded, like shapes and countours.

thats angularity, my friend

Also I'm beginning to think the result of a rating should not be a set score but a distribution a a scale

they are in place, but rather for tiers, than for the overall outcome, making it WAY more precise in itself already.

 

[Tensor]

will do if i remember

thats angularity, my friend

I'd rather say angularity is a part of it, I don't feel like angularity captures all of it and well enough.

they are in place, but rather for tiers, than for the overall outcome, making it WAY more precise in itself already.

i know
But isn't the outcome not just an average of all peoples rating?
By making it a distribution you would get more info and could maybe also incorporate things like phenotypic differences.

 

[Nerian]

There are so many problems with it. We deserve a better science of looks.

 

[zemult]

You should tag actual good raters as well
@Klasik616 @zemult @greycel @Tigermoggerlol @KT-34

Yeah I've been of this opinion for a while.

Our monkey brains simply cannot completely deterministically explain why beauty is what it is.

Depending on the perspective of the rater, everyone will give different answers.

That's why normies believe so strongly in the misconception that beauty is in the eye of the beholder. Although they're wrong and beauty is objective, we just cannot pinpoint exactly what makes beauty, because of how complex it is, which is what leads them to this conclusion. We only have abstract ideas of beauty, hence the ratios, which are usually backed by research (unlike FaceIQ's new ratios he made up, jfl).

A good rater can almost always get an idea where a person is at though, whether they're ugly, average, or GL. There's a point where someone's ratios overwhelmingly fall outside of the ideal and most of all they just fail the eye test. They don't look harmonious. Same with whether someone is good looking or not. Their ratios are usually ideal and they just look harmonious.

 

[monecel]

Nigga has to write in rainbows to get people to read his threads

 

[KKKuroiso]

@KKKuroiso @shanevasily @lurking truecel @Aryan Incel

Agree

 

[AscendingHero]

A fundamental flaw within face ratings

The face rating system that is commonly used assesses:

- Harmony
- Angularity
- Dimorphism
- Features

Out of 10 and constitutes the final rating out of 10 as such:

(Harmony score/10 )*0,4
+ (angularity score/10)*0,2
+ (Dimorphism score/10)*0,2
+ (Features score/10)*0,2

(this may be off by a little bit feom what is used right now but is close enough and won't be a problem for the argument)

Now as for the individual categories there seems to be plenty of nuance going into constituting the scores out of 10.

Spoiler: Examples

For example all the measurements going into the harmony score, which all are (or should at least be) backed by studies.

Or looking at thousands of people, assessing each persons angularity and assuming a normal distrubution and creating a scale with indicators with that.

Same with dimorphism and features (where one can even go as far as rank every feature by itself first)

Now why is it that those different categories are weighted as such?

Well the aswer is:

There does not seem to be a good reason at all!

It's just because people thought it would make sense!

While we require studies made of specific facial measurements to introduce them to our formula to costitute the harmony score (that only constitutes 40% of the harmony score)

There is not absolutely nothing indicating that it should be done this way!

(Or at least I have not seen or heard of any, although I have not really looked into it. Please correct me if I'm wrong and just
babbling nonsese)

So how can it be EVERYONE is accepting it, using it and not even questioning it

I mean sure it is somewhat sensible but just assuming that harmony has double the weight of the other categories and the categories that are not harmony all hold the same weight is
completely arbitrary
(Even if lets say its slightly different there is no actual justification)


This should not be condoned in a space where people claim to perform completely objective face ratings, which are comoletely based on studies and "math"

And this does not even include that the division into the 4 different specific categories needs to be justified

(Now again please correct me if there is justification. I might just be a dumbass rambling at 3am )

Anyways thanks for reading.
While this might have sounded a little harsh I would seriously like to know info about it and please give your opinion on it .

Decent base idea but if could get even more detailed