Yes, the Toplists are Broken

[Bazzle]

I’d like to understand what stochastic process leads to virtually every leading story in LS having exactly the same (to 2 d.p.) rating of 4.84. Very happy to entertain theories as to how this might happen by random chance. Please educate me.

Without the behind the scenes data, we can't know. BUT I expect its partially fluke of x thousand votes.

 

[RocketGrunt]

Whereas to me, a person who is by nature suspicious of people who see patterns in randomness - well, this looks very much like a pattern.

Several proposals have been floated for the reason behind the pattern. Ultimately, though, we're all just floating hypotheses. You have your favourite and I have mine, and unless things change at the structural level on the site to the tune of exposing actual voting data, at best we will be able to assemble circumstantial evidence.

Yes. It's very obviously a pattern.

Of course, there are more things going on. There are in fact, several patterns.

The targeted downvots that EmilyMiller is reporting are obviously different than the triggered downvotes that happen when cream stubbornly rises to the top on the ironically named toplists. But the METHOD looks very much the same. There is a criteria to trigger the downvotes and the downvotes come in rapidly and consistently enough that the process is clearly automated.

The difference is in criteria. EmilyMiller is reporting repeat hammerings on stories when they bubble up to 4.5. That's different from the toplist flattener that stops targeting stories when they fall to 4.84 or below. But the detection mechanism of finding stories with the offending rating and the mechanism of avoiding the site's systems for finding inauthentic votes are likely the same.

It may well be the same script, simply set to different target parameters.

 

[BobbyBrandt]

I’d like to understand what stochastic process leads to virtually every leading story in LS having exactly the same (to 2 d.p.) rating of 4.84. Very happy to entertain theories as to how this might happen by random chance. Please educate me.

A reasonable question that transcends a single category.

If you look at only the top list stories, it is easy I suppose to view them as the target. However, if you look at the fact that stories not on the top lists, or far down on them are also seeing the "leveling", it's harder to accept anything other than another glitch in the site as the most likely cause.

For whatever reason, intentional or not, there is currently a system-wide threshold limit on scores that keeps them at 4.84 or lower once they have 1000 votes. Their position on a top list is not a criteria for this adjustment.

How could this threshold be used to the advantage of someone? They couldn't get their story(ies) rated higher. Why do it?

 

[EmilyMiller]

BUT I expect its partially fluke of x thousand votes.

That’s not how numbers behave unless the process is non-stochastic. Thousands of votes are less likely to lead to such a statistical anomaly than hundreds. It’s called the law of large numbers. As the sample size increases, it’s expetation asymptomatically approaches the population expectation. This is NOT what we see here.

 

[RocketGrunt]

Without the behind the scenes data, we can't know. BUT I expect its partially fluke of x thousand votes.

That is absolutely not what it is. Large numbers of stories with large numbers of votes have Gaussian distributions. They don't look anything like this.

Not only is the big crunch at 4.84 unnatural, but it's unnatural for it to remain like that. With thousands of votes being cast, stories would pop up above 4.84 all the time. And they do. Keeping the number of stories on the Lesbian Sex toplist rated above 2.84 to less than ten (as of the moment I am writing this, there are only 6) takes constant effort. Every day stories pop up, and every day downvotes are sent to the specific stories that do to push them back down again.

Entropy naturally increases. Even if every story on the toplist "deserved" to get precisely 84 5* ratings for every 16 4* ratings, the reality of stochastic processes would give you a stratified distribution that would get noisier over time. If entropy is not increasing in a system, that is proof that there is a counter-entropic force. In this case, the counter-entropic force is triggered, targeted troll voting. This has been demonstrated, and the only people who pretend to not believe this are unconvincible.

 

[Bazzle]

That is absolutely not what it is. Large numbers of stories with large numbers of votes have Gaussian distributions. They don't look anything like this.

Not only is the big crunch at 4.84 unnatural, but it's unnatural for it to remain like that. With thousands of votes being cast, stories would pop up above 4.84 all the time. And they do. Keeping the number of stories on the Lesbian Sex toplist rated above 2.84 to less than ten (as of the moment I am writing this, there are only 6) takes constant effort. Every day stories pop up, and every day downvotes are sent to the specific stories that do to push them back down again.

Entropy naturally increases. Even if every story on the toplist "deserved" to get precisely 84 5* ratings for every 16 4* ratings, the reality of stochastic processes would give you a stratified distribution that would get noisier over time. If entropy is not increasing in a system, that is proof that there is a counter-entropic force. In this case, the counter-entropic force is triggered, targeted troll voting. This has been demonstrated, and the only people who pretend to not believe this are unconvincible.

From my side of things, the only story that ever came near to the top list was my 'Spick and Span Only with a fan' up to 90 ish votes it was at 4.84, from 100 to its now 196 votes it has resolutely stayed at 4.78. It did not move one iota as the votes clicked up. I contemplated messaging Manu to see if it was a fault, how could I get 95 votes and the score stay the same?

So I agree something is at play, but I assumed it was volume of 1, 3's and 5's.

 

[Raybies93]

Who cares?

That was a complete waste of time.

 

[EmilyMiller]

If you look at only the top list stories, it is easy I suppose to view them as the target. However, if you look at the fact that stories not on the top lists, or far down on them are also seeing the "leveling", it's harder to accept anything other than another glitch in the site as the most likely cause.

I said I’m not going to get into an argument about this and it’s kinda pointless. But what you say doesn’t address my question, it’s just waving your hands.

In a normal (no pun intended) distribution of 1 - 5 votes with an overlay of 5 being the most common - you would not expect so many stories to have exactly the same ratings. It’s statistically very, very unlikely.

Add to this the widely observed phenomenon that no sooner does a story reach 4.85 it gets downvoted and we have additional weight added to the idea of human action.

I’m not claiming I know precisely why this happens, but it’s irrefutable that something beyond random chance is happening.

And I see no reason why the site benefits beyond Rube Goldberg reasoning. And why does it seem to happen in some categories and not others if it is a site wide thing impacting all stories? LS is entirely clear cut. Unless the site has a special algorithm for that category, what else beyond human actions could be the cause?

 

[EmilyMiller]

From my side of things, the only story that ever came near to the top list was my 'Spick and Span Only with a fan' up to 90 ish votes it was at 4.84, from 100 to its now 196 votes it has resolutely stayed at 4.78. It did not move one iota as the votes clicked up. I contemplated messaging Manu to see if it was a fault, how could I get 95 votes and the score stay the same?

So I agree something is at play, but I assumed it was volume of 1, 3's and 5's.

Ratings should stabilize over time. With 1,000 votes cast the impact of the next vote becomes less.

Which actually makes it more weird that any LS story breaching 4.84 suddenly changes direction rapidly. This should not normally happen in such a predictable manner.

 

[RocketGrunt]

From my side of things, the only story that ever came near to the top list was my 'Spick and Span Only with a fan' up to 90 ish votes it was at 4.84, from 100 to its now 196 votes it has resolutely stayed at 4.78. It did not move one iota as the votes clicked up. I contemplated messaging Manu to see if it was a fault, how could I get 95 votes and the score stay the same?

So I agree something is at play, but I assumed it was volume of 1, 3's and 5's.

When stories have high ratings, they are more susceptible to downvotes than upvotes.

What you're looking at is what happens when most of your votes are 5 and a few of your votes are troll hammers. Specifically, you got a total of 27 "downvotes" among your 96 votes. But while a 4* is a downvote, a * is two, a 2* is three, and a 1* is counted as four downvotes.

Assuming 4.84 is your expected value, because that was the rating you had before being exposed to the toplist trollsript, we'd expect you to get about 15 4* reviews in 96 votes. But added to that were about three 1*, or more likely two pairs of 2* ratings.

So probably right around the hundred mark, you got a run of 5* ratings that pushed you briefly into the top list, exposing you to the downvote bot. The downvote bot triggered on you twice, giving an extra 12 stars of down voting, which dropped your overall rating from 4.85 or 4.86 down to 4.78. And now the story is unlikely to get a lot more voting in any direction because it's no longer a new story or on a toplist or anything, meaning that many fewer people will read it or vote on it.

As to why it looked like the rating did not move, the ratings on most pages don't update very often, so there was probably a lot of movement in the day that didn't catch up to being displayed until the real votes and the troll votes had achieved a new equilibrium. This delay in score updates might also be why you got double tapped twice rather than once.

 

[ThatNewGuy]

I said I’m not going to get into an argument about this and it’s kinda pointless. But

What's the old saying? "You can take the girl out of the dick-swinging party..."

 

[EmilyMiller]

What's the old saying? "You can take the girl out of the dick-swinging party..."

I should maybe have said, I’m not gonna continue to debate after this post, but…

And my multi-dick days are now long gone

 

[EmilyMiller]

The following is something I posted on the Valentine’s thread. It’s not directly pertinent to top lists vote suppression, but indicative of the malicious human actions that are very common here:

—

To prepare writers who are entering their first competition here (and are really happy with their current rating) for what can happen in the closing days, here is some actual data from one of my old comp entries. The data is from a story in a low frequency voting category, one in which I could count every vote and determine its value precisely.

Number of non-five star votes in the first eighty votes: 7

Number of non-five star votes in the last six votes before the winners were announced: 3

Using the binomial distribution formula to assess the probability of the second set of events given the first we get 1.85%.

That means the likelihood of the voting pattern in the last six votes being purely due to chance is less than two in a hundred.

People will go on about followers voting early and “regular people” voting later. But, even in a comp, the vast majority of votes are early on. And if you think this factor is a better explanation, well then .

So, new authors, if you suddenly see you much cherished story have its rating dip in the last few days of voting, this is why, and you are not alone. It’s endemic.

UPDATE: the aggregate votes on this story didn’t break 100 during the comp, so the well-documented top lists-related vote manipulation was not a factor

 

[RocketGrunt]

Well, let's watch these then.

View attachment 2595749

At some point this year, a couple of these are going to break 100 votes. At that point, they're going to get hit just like Ch. 04 & 09 did. The toplists are undeniably a hotbed of trolling. ( And they're not going to uniformly drop to 4.83 either, because it's uncoordinated humans doing it, and not a script )

Update:

One Whore's Town Ch, 03 got one more vote (presumably 5*) and appeared briefly at the top of the toplist with a 4.91 rating. It then departed the toplist a few hours later and has a rating of 4.77.

I can't see the exact vote tally, but that is consistent with having been zapped by four 1* ratings within hours of being exposed to the toplist. Probably within minutes.

A story published in 2024 sat comfortably at 4.91 with 99 ratings for however many months, and exposure to the toplist gave it four 1* ratings from people who obviously did not read or rate any other chapter in the series. This does match the predictions of the people who say there's a downvote bot scripted to target unprotected stories that come onto the toplist above 4.84. It matches that exactly. It is completely incompatible with any of the other explanations that have been put forward.

Except the idea that there are multiple trolls who are doing it manually by collectively downvoting things consistently enough that their actions are indistinguishable from a python script. That obviously cannot be discounted, since it is by definition indistinguishable from an automated trolling process.

 

[Voboy]

Update:

One Whore's Town Ch, 03 got one more vote (presumably 5*) and appeared briefly at the top of the toplist with a 4.91 rating. It then departed the toplist a few hours later and has a rating of 4.77.

I can't see the exact vote tally, but that is consistent with having been zapped by four 1* ratings within hours of being exposed to the toplist. Probably within minutes.

A story published in 2024 sat comfortably at 4.91 with 99 ratings for however many months, and exposure to the toplist gave it four 1* ratings from people who obviously did not read or rate any other chapter in the series. This does match the predictions of the people who say there's a downvote bot scripted to target unprotected stories that come onto the toplist above 4.84. It matches that exactly. It is completely incompatible with any of the other explanations that have been put forward.

Except the idea that there are multiple trolls who are doing it manually by collectively downvoting things consistently enough that their actions are indistinguishable from a python script. That obviously cannot be discounted, since it is by definition indistinguishable from an automated trolling process.

Useful data. Wicked disheartening.

 

[Statius]

Competing site trying to discourage readers and writers? Otherwise, why?

For the record I have little to no skin in this, just postulating.

 

[Voboy]

Competing site trying to discourage readers and writers? Otherwise, why?

I can think of many possibilities, all of them consistent with my generally negative experience of collective human nature in anonymous environments.

 

[RoyalAuthor]

I had a story a month ago hit 4.89 on its hundredth vote. Not only did it immediately start getting the 2s and 3s, it got 2 or 3 votes a night for three weeks driving it down to 4.45 before I finally froze the voting.

I have a story that's at 4.89 with 98 votes right now. Can't wait to see what happens tomorrow.

Here's another data point: crossed into the all time/12 month toplist for group sex at 4.88 with 101 votes last night. Woke up this morning with a 4.78 after 104 votes.

Nearly every story I write hits the top list for a few hours and then immediately gets beaten down. Almost all of them finish in the high 4.7s or higher, with a bunch of them at 4.84. Occasionally one will rise to 4.85 from steady voting - without fail it's below 4.8 by the next morning with 2 or 3 more votes, even if the story is months old and only getting a vote or two a week at that point.

I'd buy the "larger reader pool" argument if it wasn't always 2 or 3 extremely negative votes immediately, never a smattering of different scores that drag it down. You can set your watch by it.

 

[EmilyMiller]

Here's another data point: crossed into the all time/12 month toplist for group sex at 4.88 with 101 votes last night. Woke up this morning with a 4.78 after 104 votes.

Nearly every story I write hits the top list for a few hours and then immediately gets beaten down. Almost all of them finish in the high 4.7s or higher, with a bunch of them at 4.84. Occasionally one will rise to 4.85 from steady voting - without fail it's below 4.8 by the next morning with 2 or 3 more votes, even if the story is months old and only getting a vote or two a week at that point.

I'd buy the "larger reader pool" argument if it wasn't always 2 or 3 extremely negative votes immediately, never a smattering of different scores that drag it down. You can set your watch by it.

There are one off anedotes, which may not mean much. Then there is author after author after author having exactly the same experience; sometimes on multiple stories.

This is NOT something that is a result of ‘adjustments’ impacting all stories. It is very specifically related to entry into the all-time top list for certain categories.

To claim it’s anything other than human action (probably aided by automation) is to adopt this stance:

 

[BobbyBrandt]

There are one off anedotes, which may not mean much. Then there is author after author after author having exactly the same experience; sometimes on multiple stories.

This is NOT something that is a result of ‘adjustments’ impacting all stories. It is very specifically related to entry into the all-time top list for certain categories.

To claim it’s anything other than human action (probably aided by automation) is to adopt this stance:

View attachment 2595975

With all due respect, did you look at any other categories besides LS?

Every category top list that I see now has the top rated story no higher than 4.85, while many are at 4.84. If you do as I did, and track all my own stories over the past year, you would see that the vote adjustments are not limited to stories only in the top lists. I currently have eight stories "stuck" at 4.84.

So, I'll ask the question again, "Why?"

Why would some mythical human actor go to the trouble of establishing this top score threshold for all categories? How could it possibly benefit them?

A crime without any motivation is hard to prove.

I don't ignore that the means and opportunity exist from a technological perspective, but please explain the motive to me in a way that make sense.

"Because they can," is a cop-out.

 

[SimonDoom]

I'm always struck by the enormous amount of energy and distress that goes into noodling over this subject of downvoting and how the site handles it.

Let's assume all of it is true.

Who cares? Why care?

How does it stop you from enjoying your experience as a Literotica author? It doesn't, unless you get caught up in caring more about lists and scores than you should. You don't have to care about any of that stuff to have fun as a Literotica author.

It's a choice. Why choose being upset? It's an imperfect system, but you can still have fun and get creative satisfaction from publishing stories and connecting with authors who like what you do. And that's what matters, not these silly scores and toplists.

 

[pink_silk_glove]

Why would some mythical human actor go to the trouble of establishing this top score threshold for all categories? How could it possibly benefit them?

Because he wants to control who is or who isn't on top. Simple.

It would take a petty narcissistic asshole to do that. There are plenty of those just in the AH here, and the AH is but a small cross section of writers. So do the math, it's pretty much impossible to NOT find someone out there willing to do this.

Sure, it sucks, but so what? If we can come up with a solution, great, but none of the solutions that I've seen proposed fix anything.

 

[Voboy]

I'm always struck by the enormous amount of energy and distress that goes into noodling over this subject of downvoting and how the site handles it.

Let's assume all of it is true.

Who cares? Why care?

How does it stop you from enjoying your experience as a Literotica author? It doesn't, unless you get caught up in caring more about lists and scores than you should. You don't have to care about any of that stuff to have fun as a Literotica author.

It's a choice. Why choose being upset? It's an imperfect system, but you can still have fun and get creative satisfaction from publishing stories and connecting with authors who like what you do. And that's what matters, not these silly scores and toplists.

This, to me, is the crux.

It’s easy to convince me that shenanigans are occurring. It’s just as easy to convince me that the site doesn’t care. So… where does that leave any of us?

A million of these threads, and the answer is the same: if this is something that bothers you, then you’ll need to find a way to stop it from bothering you.

That’s up to each of us. I don’t find it difficult to do, but I’m aware other people have a hard time letting this go.

 

[EmilyMiller]

This, to me, is the crux.

It’s easy to convince me that shenanigans are occurring. It’s just as easy to convince me that the site doesn’t care. So… where does that leave any of us?

A million of these threads, and the answer is the same: if this is something that bothers you, then you’ll need to find a way to stop it from bothering you.

That’s up to each of us. I don’t find it difficult to do, but I’m aware other people have a hard time letting this go.

I’m with you. My choice was to stop publishing here. It’s either that or find a way to put up with things. There is no third way.

But the issue as I see it is when people deny that anything is happening. Of course the site is not going to lift a finger. But telling authors who have been clearly impacted that they are imagining it is not cool. What happened to a bit of empathy and solidarity?

 

[pink_silk_glove]

But telling authors who have been clearly impacted that they are imagining it is not cool. What happened to a bit of empathy and solidarity?

Call me callous but I have no empathy for the butthurt nor the bruised egos.