Why shouldn't we rate with percentile?

[Patient A]

Why not just make conversions of percentile to ratings? So when we determine the rating we are using the original method you mentioned (using the dataset), we use percentiles. For the final rating, though, we switch it back to the 0-10 scale. A good dataset would be your highschool yearbook btw if your town is large enough. Here's an idea:

Throw out oldcels, people not of your race, and excessively fat people (30% for men) from the dataset as these skew the data.
2% = 3/10
16% = 4/10
50% = 5/10
84% = 6/10
98% = 7/10

If you observe that someone is less attractive than the majority (but not an overwhelming majority) of young men, then you can say that he is a 4/10 without discussing percentiles to others and sounding autistic.

What the fuck are you talking about retard

 

[Deleted member 15827]

Why not just make conversions of percentile to ratings? So when we determine the rating we are using the original method you mentioned (using the dataset), we use percentiles. For the final rating, though, we switch it back to the 0-10 scale. A good dataset would be your highschool yearbook btw if your town is large enough. Here's an idea:

Throw out oldcels, people not of your race, and excessively fat people (30% for men) from the dataset as these skew the data.
2% = 3/10
16% = 4/10
50% = 5/10
84% = 6/10
98% = 7/10

If you observe that someone is less attractive than the majority (but not an overwhelming majority) of young men, then you can say that he is a 4/10 without discussing percentiles to others and sounding autistic.

This forum is already autistic with psl ratings percentiles wouldn't be a different story sadly

 

[untouchable_coper]

What the fuck are you talking about retard

Percentiles seem to be more accurate than PSL/1-10, but are harder to converse about. Using percentiles to base the ratings off of, though, would solve this problem.

 

[Deleted member 15827]

Percentiles seem to be more accurate than PSL/1-10, but are harder to converse about. Using percentiles to base the ratings off of, though, would solve this problem.

Thing is no one knows how avg Indian or middle East pop look like so it'll still be rated with cumskin population in mind, very hard to adjust for these things. I think if all the curries made themselves known here we can form a group chat where we rate each other in percentiles

@inferiorpispot234 @subhuman pakicel @SubhumanCurrycel @Kmscurry @currylightskin @Curry Suicide @Madhate @pianoboy123 @VicMackey @volcelfatcel

 

[currylightskin]

Thing is no one knows how avg Indian or middle East pop look like so it'll still be rated with cumskin population in mind, very hard to adjust for these things. I think if all the curries made themselves known here we can form a group chat where we rate each other in percentiles

@inferiorpispot234 @subhuman pakicel @SubhumanCurrycel @Kmscurry @currylightskin @Curry Suicide @Madhate @pianoboy123 @VicMackey @volcelfatcel

True

 

[Deleted member 15827]

True

U know how to create?

 

[Deleted member 16346]

Thing is no one knows how avg Indian or middle East pop look like so it'll still be rated with cumskin population in mind, very hard to adjust for these things. I think if all the curries made themselves known here we can form a group chat where we rate each other in percentiles

@inferiorpispot234 @subhuman pakicel @SubhumanCurrycel @Kmscurry @currylightskin @Curry Suicide @Madhate @pianoboy123 @VicMackey @volcelfatcel

Why would anyone wanna be in a group chat where all we do is rate you, pianoboy and currylightskin over and over again tbh

 

[Deleted member 15827]

Why would anyone wanna be in a group chat where all we do is rate you, pianoboy and currylightskin over and over again tbh

Doesn't that sound fun

 

[thecel]

Why not just make conversions of percentile to ratings? So when we determine the rating we are using the original method you mentioned (using the dataset), we use percentiles. For the final rating, though, we switch it back to the 0-10 scale. A good dataset would be your highschool yearbook btw if your town is large enough. Here's an idea:

Throw out oldcels, people not of your race, and excessively fat people (30% for men) from the dataset as these skew the data.
2% = 3/10
16% = 4/10
50% = 5/10
84% = 6/10
98% = 7/10

If you observe that someone is less attractive than the majority (but not an overwhelming majority) of young men, then you can say that he is a 4/10 without discussing percentiles to others and sounding autistic.

the standard deviation does not equal 1

 

[untouchable_coper]

the standard deviation does not equal 1

It was just an idea. Do you think that the distribution of attractiveness is normal? If so, what is the standard deviation in your opinion?