I know Lazeez hates discussions about the scoring system, but I stumbled upon this article today that parrots what he's been saying for years:
https://hbr.org/2019/07/the-problems-with-5-star-rating-systems-and-how-to-fix-them
Basically all rating systems suffer from raters' bias, grade inflation and bunching, etc...
This is why I've advocated a simple 'thumbs up' or 'thumbs down' system. And show the raw numbers as well as a ratio/percentage. Done and dusted. At this point, it would, IMHO, convey better information without trying to parse the difference between a rating of '8' and a rating of '9'.
without trying to parse the difference between a rating of '8' and a rating of '9'.
But you don't. You're choosing between "very good" and "great."
But you don't. You're choosing between "very good" and "great."
You are, so is every other voter. The problem is WHY voters award the score.
Reader A rates it a 9 because it is beautifully crafted without spelling or grammatical errors.
Reader B rates it an 8 because she had massive tits, would be a 10 if he had knocked up her daughter.
Reader C rates it an 8 because it's a good story but included spanking and BDSM is a squick.
The problem is WHY voters award the score
No scoring system will fix subjectiveness. In your example, Reader B might give it a thumbs up because of the massive tits whereas Reader A might give it a thumbs down because he doesn't like big knockers.
No scoring system will fix subjectiveness.
Agreed, given that fact it really does not matter what system is used.
Personally my only objection is the 'algorithm' but I'm not expecting that to change anytime ever, so as always, threads on scoring are pretty much pointless.
Reader D rates it a 10 because, although the story kinda sucks, they like the author and want to encourage them to write more in the hope the next story marks a return to what they like.
The unfathomable whimsicality of voting reasons ...
AJ
a simple 'thumbs up' or 'thumbs down' system
The thumbs up or down rating system also has a major problem IMHO. The system is an all or nothing system that presents the rater with votes for totally acceptable and totally unacceptable. Most stories fall between those two extremes. The system fails to provide the rater with a means of discriminating between different degrees of acceptable/unacceptable.
A rating system of 1 to 5, or other values, at least offers the rater the means of indicating whether the story was basically acceptable with flaws or basically unacceptable with some merit.
The thumbs up or down rating system also has a major problem IMHO. The system is an all or nothing system that presents the rater with votes for totally acceptable and totally unacceptable. Most stories fall between those two extremes. The system fails to provide the rater with a means of discriminating between different degrees of acceptable/unacceptable.
A rating system of 1 to 5, or other values, at least offers the rater the means of indicating whether the story was basically acceptable with flaws or basically unacceptable with some merit.
A rating system of 1 to 5, or some other values,
We already have that...
1 = 0.5
2 = 1
3 = 1.5
4 = 2
5 = 2.5
6 = 3
7 = 3.5
8 = 4
9 = 4.5
10 = 5
Et voilà, the current system now fits your desire.
the current system now fits your desire.
Actually, I prefer a system where there are as many options above the median value as below. So 0-10, 1-9, 1-5, etc.
Actually, I prefer a system where there are as many options above the median value as below. So 0-10, 1-9, 1-5, etc.
Any scoring system does have this! The median is defined this way (equal number of values below and above).
The only problem with the median value of the existing system (1-10) is you can't chose it. (5.5). It has 5 values below (1, 2, 3, 4, 5) and 5 values above (6, 7, 8, 9, 10).
BTW, with those evenly spaced values (1-10) the mean is the same as the median.
HM.
mean is the same as the median.
No, it is a little more complicated, at least statistically.
Here is something I found on-line that may help. At least I don't have to confuse you with another measure of central tendency, the mode. (The most common number in the progression.)
"Are mean and median the same thing?
4 Answers
Anirvan Maiti
Anirvan Maiti, M.S. from Boston University
Updated Aug 16 2015
Lets say we have a set of 7 numbers as follows :
72, 2, 6, 32, 55, 15, 9
Mean
Mean is the average of a set of numbers.
So the mean for the above given set of numbers is :
72+2+6+32+55+15+9
7
Which gives the answer : 27.2857
Median
Median is the middle element in an ordered set.
So, if we arrange the above set in an ascending order, we get :
2, 6, 9, 15, 32, 55, 72
We see that the middle most element in this ordered set is the number 15
Therefore, 15 is the median of the set.
Difference between Mean and Median (CONCEPT)
In the above given set of ordered numbers, if I change the last number 72 to 99999, the mean becomes
2+6+9+15+32+55+99999
7
which is equal to 14302.5714
BUT
The median of this set still remains the same. Because in this set
2, 6, 9, 15, 32, 55, 99999
the middle element still remains the same, which is 15
Relation between Mean and Median
If we have a set where every number is equidistant from its predecessor (or successor) like
3, 6, 9, 12, 15
Then the mean
3+6+9+12+15
5
= 9
and the median (middle element is 9) are equal !"
No, it is a little more complicated, at least statistically.
I'm well aware of the differences between mean, median and mode.
However you didn't quote correctly.
I wrote
with those evenly spaced values (1-10) the mean is the same as the median.
English isn't my native language. I meant the same you cited:
If we have a set where every number is equidistant from its predecessor (or successor) [...] Then the mean [...] and the median [...] are equal !
HM.
Richardshagrin has a valid point, although he did not stress it.
Median
The median is the value separating the higher half of a data sample, a population, or a probability distribution, from the lower half. In simple terms, it may be thought of as the "middle" value of a data set.
https://en.wikipedia.org/wiki/Median
The current rating system is a data set consisting of 10 values. The mean (average) of the data set is 5.5 as you stated. However the median value must be a value that is listed in the data set and the median value cannot be 5.5 because 5.5 in not a value in the data set.
However the median value must be a value that is listed in the data set and the median value cannot be 5.5 because 5.5 in not a value in the data set.
This is not true. From your own cite to the Wikipedia entry on median.
If there is an even number of observations, then there is no single middle value; the median is then usually defined to be the mean of the two middle values.[1][2] For example, in the data set
1, 2, 3, 4, 5, 6, 8, 9
the median is the mean of the middle two numbers: this is ( 4 + 5 ) / 2 {displaystyle (4+5)/2} {displaystyle (4+5)/2}, which is 4.5 {displaystyle 4.5} {displaystyle 4.5}. (In more technical terms, this interprets the median as the fully trimmed mid-range).
I consider my hand as being slapped for not reading all of what I cited. :(
In my original statement, it was my intent to define a rating system that had an odd number of values in its data set of selectable values.
is you can't chose it
So I didn't specifically state the median value has to be selectable, but that is what I meant by median value.
sexstories.com has simple thumbs-up and thumps-down voting, and I'm not sure its any better.
As I've told everyone too many times, collaborative filtering is what we need. Its really simple, and its how pretty much every online shopping site comes with recommendations for other products when you view something.
So, let people score stories, and then suggest to people the stories that other people who rated similiarly to they do also rated highly.
The maths is hard to understand but easy to implement, but there are a gazillion python implementations of slope-one or some similar collaborative filtering algorithm so these days, nobody has to actually implement it.
Just feed the scoring history in, and hey-presto, problem solved!
As I've told everyone too many times, collaborative filtering is what we need.
More accurately, it's what YOU think we need.
Opinions vary.
