#### Gymcelled

##### Genetically shackled to hell

**★★★★★**

- Joined
- Jul 15, 2019

- Posts
- 11,024

- Online
- 120d 15h 54m

This will probably be my most autistic post yet but i think it's seriously needed.

tl;dr this post will mathematically explain to you how hard it is to compensate for a singular atrocious flaw

It's possible for both manlets and ugly tallfags to exist simultaneously because being extremely bad at one of the two features (face or height) will induce a larger penalty on your looks score

Don't think in simple arithmetic averages

The point isn't to discuss which one is more important but rather how they interact with each other

On one hand you have really short guys with decent faces looking at their life experiences and concluding that face is NOT more important than height, because it hasn't saved them. Ugly tallfags do the same with their height. Each member of the two camps is convinced that the other one is full of bullshit.

People often say that height is the "looks multiplier". I disagree, I think it's more complicated than that.

I will attempt to explain how EVEN IF face and height have the EXACT SAME worth, we can still prove with simple mathematics that it's possible to fail to compensate with the other feature.

(we're trying to simplify the calculations, if you disagree: keep in mind this equality only strengthens what i'm about to show)

At the end I show you that this still works even if you attribute a different worth to face and height, and you're free to expand the model by adding in more variables.

First a quick analogy to explain what this post is about

Imagine there’s a smartphone that comes out and has to be rated. It’s amazing at everything: fast, great battery, amazing software, good camera, nice speakers, looks aesthetic … but it weighs 20lbs (or 10kg). Would anyone want that phone? Being so bad at that one singular thing ruins everything and makes it worthless as a phone. Yet if you were to rate all of its feature, add all the ratings up and average it, it would look like an amazing phone right?

Averaging things with a simple arithmetic mean is not always relevant.

If you're extremely short OR extremely ugly but you're amazing in every other area of life ... YOU'RE THE 20LBS PHONE, YOU'RE A FAILED PRODUCT THAT NO ONE WANTS

The arithmetic mean doesn’t care about the variance (spread) of your data. It doesn’t take into account how a god awful feature might ruin everything.

The geometric and the harmonic mean are better suited for rating looks because they impose a penalty based on the spread of the face and height scores. If you’re really bad at either face or height it’ll take that into account and compensating with the other will be much harder.

The formulas for geometric and harmonic means are given below

By only taking into account face and height and by attributing the same worth (ie we assume that face and height are equally as important), the formulas simply become sqrt(f*h) and (2*f*h)/(f+h) for the geometric and the harmonic mean respectively (where f and h are the face and height scores, from 0 to 10).

As you can see below, I make face go from 0 to 5 and height from 10 to 5. This would be an ugly guy trying to compensate with height. If we simply apply the arithmetic mean we get 5/10 in looks regardless of the combination. As if being a 4 and 6/10 in face and height is the same as a 1 and 9/10.

Meanwhile, the geometric and harmonic mean will impose a penalty on the total looks score when either face or height gets very low. So a guy with a REALLY BAD face will still have a bad looks score, just like a turbomanlet would if he tried to compensate with face.

Personally I prefer the harmonic mean because it’s stricter than the geometric mean. As you can see, someone who scores a 2 and 8/10 in face and height (or the reverse case) will only be a 3/10 (3.2 to be exact).

This is how you mathematically explain both turbomanlets AND ugly tallfags being a thing. Even if face and height were equally as strong you wouldn’t be able to compensate perfectly with the other one simply because of the 20lbs smartphone effect. Your flaw is just too big to be ignored, it ruins everything.

With that we can generate curves for different height scores and make face vary.

So with H=2 for instance (turbomanlet) you can see that increases in face barely help the shortcel at all.

Same graph but with less curves and zoomed in to show.

One more example comparing arithmetic and harmonic mean with height = 7/10 and a varying face.

So let’s say you disagree with face and height being 50/50. You think one or the other is more important. Let’s say 60/40 or 30/70 for instance. This model still works, you just have to apply the formula below.

You can also add more variables (race, frame, whatever).

I did some quick curves with 60/40 and 70/30. You can investigate further if you're curious. If you think the spreading penalty is too big just use the geometric mean instead and apply weights to the formula.

Lastly I think this penalty I talked about only starts kicking below 5/10 and only becomes significant below 4/10. If someone were to score above 5/10 in both face and height then forget the harmonic mean, the arithmetic mean should be more accurate because you don't have that big failo and "ew 20lbs phone" effect.

@Selinity @soymonkcel @ionlycopenow @Edmund_Kemper @your personality

@SergeantIncel @Master @mental_out

@ReturnOfSaddam How is that for autism?

tl;dr this post will mathematically explain to you how hard it is to compensate for a singular atrocious flaw

It's possible for both manlets and ugly tallfags to exist simultaneously because being extremely bad at one of the two features (face or height) will induce a larger penalty on your looks score

Don't think in simple arithmetic averages

The point isn't to discuss which one is more important but rather how they interact with each other

- The manlet vs ugly tallfag conundrum

On one hand you have really short guys with decent faces looking at their life experiences and concluding that face is NOT more important than height, because it hasn't saved them. Ugly tallfags do the same with their height. Each member of the two camps is convinced that the other one is full of bullshit.

People often say that height is the "looks multiplier". I disagree, I think it's more complicated than that.

I will attempt to explain how EVEN IF face and height have the EXACT SAME worth, we can still prove with simple mathematics that it's possible to fail to compensate with the other feature.

**Hypothesis 1:**Face and height have the exact same worth (50/50).(we're trying to simplify the calculations, if you disagree: keep in mind this equality only strengthens what i'm about to show)

**Hypothesis 2:**No other factor come into play when it comes to your looks score. (again we're doing this to simply the examples)At the end I show you that this still works even if you attribute a different worth to face and height, and you're free to expand the model by adding in more variables.

First a quick analogy to explain what this post is about

- 20lbs smartphone analogy, why linear arithmetic thinking is wrong

Imagine there’s a smartphone that comes out and has to be rated. It’s amazing at everything: fast, great battery, amazing software, good camera, nice speakers, looks aesthetic … but it weighs 20lbs (or 10kg). Would anyone want that phone? Being so bad at that one singular thing ruins everything and makes it worthless as a phone. Yet if you were to rate all of its feature, add all the ratings up and average it, it would look like an amazing phone right?

Averaging things with a simple arithmetic mean is not always relevant.

If you're extremely short OR extremely ugly but you're amazing in every other area of life ... YOU'RE THE 20LBS PHONE, YOU'RE A FAILED PRODUCT THAT NO ONE WANTS

- Using other means to evaluate looks

The arithmetic mean doesn’t care about the variance (spread) of your data. It doesn’t take into account how a god awful feature might ruin everything.

The geometric and the harmonic mean are better suited for rating looks because they impose a penalty based on the spread of the face and height scores. If you’re really bad at either face or height it’ll take that into account and compensating with the other will be much harder.

The formulas for geometric and harmonic means are given below

By only taking into account face and height and by attributing the same worth (ie we assume that face and height are equally as important), the formulas simply become sqrt(f*h) and (2*f*h)/(f+h) for the geometric and the harmonic mean respectively (where f and h are the face and height scores, from 0 to 10).

As you can see below, I make face go from 0 to 5 and height from 10 to 5. This would be an ugly guy trying to compensate with height. If we simply apply the arithmetic mean we get 5/10 in looks regardless of the combination. As if being a 4 and 6/10 in face and height is the same as a 1 and 9/10.

Meanwhile, the geometric and harmonic mean will impose a penalty on the total looks score when either face or height gets very low. So a guy with a REALLY BAD face will still have a bad looks score, just like a turbomanlet would if he tried to compensate with face.

Personally I prefer the harmonic mean because it’s stricter than the geometric mean. As you can see, someone who scores a 2 and 8/10 in face and height (or the reverse case) will only be a 3/10 (3.2 to be exact).

This is how you mathematically explain both turbomanlets AND ugly tallfags being a thing. Even if face and height were equally as strong you wouldn’t be able to compensate perfectly with the other one simply because of the 20lbs smartphone effect. Your flaw is just too big to be ignored, it ruins everything.

With that we can generate curves for different height scores and make face vary.

So with H=2 for instance (turbomanlet) you can see that increases in face barely help the shortcel at all.

Same graph but with less curves and zoomed in to show.

One more example comparing arithmetic and harmonic mean with height = 7/10 and a varying face.

- What if face and height don't have the same worth? (Face > Height or Face < Height)

So let’s say you disagree with face and height being 50/50. You think one or the other is more important. Let’s say 60/40 or 30/70 for instance. This model still works, you just have to apply the formula below.

You can also add more variables (race, frame, whatever).

I did some quick curves with 60/40 and 70/30. You can investigate further if you're curious. If you think the spreading penalty is too big just use the geometric mean instead and apply weights to the formula.

- DISCLAIMER

Lastly I think this penalty I talked about only starts kicking below 5/10 and only becomes significant below 4/10. If someone were to score above 5/10 in both face and height then forget the harmonic mean, the arithmetic mean should be more accurate because you don't have that big failo and "ew 20lbs phone" effect.

@Selinity @soymonkcel @ionlycopenow @Edmund_Kemper @your personality

@SergeantIncel @Master @mental_out

@ReturnOfSaddam How is that for autism?

Last edited: