Forum: Bug Report and Feature Requests

It does have a slightly Designed by Committee feel to it :)

Yes a weekly vote count would be a useful thing wouldn't it.

And if you did away with the current voting system

Lazeez has already stated that he has no intention of changing the voting system. I suspect that he will only make a change to it when he finds a problem with it.

Lazeez has already stated that he has no intention of changing the voting system.

I suspect any reader suggested change would only be implemented if there was a clear cost/benefit advantage. And the chance of that happening is somewhere between nought and zero because all the current possible scoring systems are no better than what he's already implemented.

AJ

Well, this is my opening point. The current system masks high quality but polarising stories, and writers can't easily tell what a vote was or even that we've got one.

There are scoring systems that don't have these problems.

There are scoring systems that don't have these problems.

Not in a single figure - just as you can't meaningfully combine means and standard deviations into a single figure. And if you supply more than one figure, you then have a disproportionate amount of effort to maintain and provide sorting etc for it. The cost/benefit is a fail, that's why Amazon/Ebay etc don't do it.

AJ

I think there's plenty of room in the system, I mean the Stats table has 10 columns. But if a story has lots of Likes that's a recommendation isn't it?

Another writer has messaged me how he has an average score with plenty of 10s but also some 1s for being contentious.

Only about 1% or something of readers vote at the moment I think? So the simple quantity of Likes would be a good indication if it's worth checking the tags.

But if a story has lots of Likes that's a recommendation isn't it?

How, from a single figure, could you discern whether a story is polarising?

Counting likes is less granular and informative than the current weighted averages system.

AJ

You would need likes and dislikes. Then you could use a like/dislike ratio.

IMO, adding a 'dislike' facility would be against the 'constructive criticism' ethos of the site. It would discourage the authors of stories with a high number of dislikes, and it would present a less positive image of the site to potential joiners. There's a reason SwampBezos etc don't have such a system.

How would existing story scores be massaged into a like/dislike system? 4 stars or fewer constitute a
'dislike', five stars or more constitute a 'like'?

AJ

Half the current rating options are negative, and we can't say they're constructive can we.

It is a question whether to offer a downvote, but then it's a further question what the system does with them. Some sites let users express themselves and then quietly ignore them, or only use them in certain metrics.

But yes a recode is pretty straightforward, with a 7 "good" and up being a Like.

Well we writers only get the average score. Readers can express more, yes, but the writer doesn't see it. Once a story has a few hundred votes you might add or subtract .1, or you might not. We might notice, or not.

In the beginning we could see all the scores, and as you say the degrees of your reaction would make a difference. Then that detail disappeared and we're left with a system that I think evolved some unintended consequences, so we may not know whether you enjoyed it at all.

It's a fair question, and yes Likes wouldn't tell readers it was polarising. What you'd get is more likes for a quality story whether it has contentious elements or not.

Also we'd get more voting, with it being simple and familiar.

I'm not proposing dislikes.

The average score is not very informative. Any rating system works a bit of course, but at the moment it's an illusion really. Readers make their choice out of 10 but then it gets mixed down, till it conveys less than a count of Likes.

The scores are so processed, I don't even know if a 6.8 means 'not bad' or 'good'. The innocent reader doesn't realise of course.

The average can be distorted by readers objecting to things in the story other than its quality, this is what I mean by polarising. They can get 1s because characters are anti religion or hunting for example.

All the scores are adjusted so that they average 6. So the 6.8 or whatever might have been what readers selected, or not.

In the beginning Lazeez showed a little chart for each story with the raw scores, but I'm guessing after lots of tearful complaints he's kept trying to save us from our insecurities. But where we've ended up might not be the best thing.

And a lot of billion dollar sites have done a lot of work on it and gone for Likes and Follows.

It says so in black and white on the site:
https://storiesonline.net/h/8/how-are-scores-calculated-and-how-does-scoring-work-in-general

It's the median, the midpoint score rather than the average, but same thing really:
"The score weighting formula figures where the story's raw score sits between this median and the extremes of 1 and 10. Then it calculates the same relative location for a median of 6.00. So if a story's raw score is equal to the raw scores median, it will end up with a score of 6. if it has a perfect 10 score, it remains at 10. If it has a raw score of 1, then it stays at 1. Stories with raw scores closest to the raw median are moved the most on the scale. Stories closer to the extremes are moved less. The raw scores median is calculated twice per day.

The weighted score gets used as the story's score for display on the site."

So, if the raw median is 7, most stories get 1 taken off their score to make it 6. Or conversely, if the raw median is 5, most storie get 1 added.

So yes as you say the ratings are an indication, but they don't actually tell us how many readers thought a story was Good. Also the average means a quality but polarising story looks average, we don't know what any recent scores were, and it's not easy to know whether there have been any recent votes at all.

You'll see that Lazeez has done a lot of work trying to make the 10-point scale work, but the manipulation arises because of the 10-point scale itself. Evidently it started trying to give writers great feedback, but now it's evolved into something unintended.

Why are you excluding posts and videos? They are equally entertainment that people enjoy or not. Can we even imagine how much research has gone into the voting system of a company like YT that's worth half a trillion dollars?

Yes here readers have a choice of 10 ratings, but nobody sees it, it just gets averaged with all the other votes. What is actually published has no granularity at all.

Readers are more likely to vote with a simple, familiar choice that conveys a basic emotion, rather than looking for a closest match among 23 words of various value judgments. It's why Likes are so universal.

The scores are so processed, I don't even know if a 6.8 means 'not bad' or 'good'

There is a simple answer to that. Read some 6.8 scoring stories and decide whether you like them or not. If you like them, then you know that's your minimum threshold. If you hate them, then you know to raise your minimum threshold past 7.

The main problem with all voting systems is the voter.

The voters' rationale for their votes is not based on the stories' merits. For example, I recall a voter stating in the forum that he always gives a story a '10' rating because the author expended a lot of time and effort to write the story. There are also a number of other reasons stated by voters that have nothing to do with the quality or merits of the story - typically rating the author or the type of story; such as, I hate polygamy stories.

Yes true, and we see how the current system has worked trying to manage voters with the sampling and normalising. And also to manage us delicate writers centring the scores to a sweet six!

Doesn't this all arise from using a 10-point scale though? If it were just a Like, or not, those issues wouldn't arise in the first place.

I would say that most of the posts in this thread are based on the assumption that voters have a valid reason for the rating they assign to a story. Unfortunately, that assumption is not true for the majority of the ratings assigned by the voters.

I believe that most voters vote against a story. One or two scenes/things in a excellent story are personally offensive so they rate the story very low. Voters don't have to justify their ratings, so you never know why the voter rated a story the way they did. Since you don't know the WHY of the voters' ratings, you don't know if the adjusted rating is valid.

The existing votes could pretty easily be recoded as Likes, Loves or null.

Second that.
'likes' have this (anti-)social media taste to them. SOL is a symbiosis of mostly amateur writers and readers, and we writers need a more differentiated feedback than the number of 'likes'.

How is an average more differentiating? My point is exactly that after a while we can't even tell what a vote was, whether they liked it or hated it.

It's a good system, and everyone is used to it. But for a reader, it doesn't tell you whether the votes are all 10s and 1s, or all 7s.