Yes, the Toplists are Broken

[FrancesScott]

If this thread was a dog, a vetenarian would have put it out of its misery by now.

 

[SimonDoom]

While Abdul Alhazred demands a sweep because the score on Necronomicon chapter 309 went from a 4.91 to a 4.89

It's eldritch.

 

[RejectReality]

Where does the 103 votes report come from? Like, obviously it's not currently correct, but presumably it is a total that existed at some time for it to be seen and recorded.

Unless your theory is that it is a report of a vote tally that never existed, it's a snapshot from a specific point in time. If it's 103 votes, that's a point in time that is between the time when it was snapshotted at 100 votes on the Toplist and when you checked in on it at 106 votes.

And if it is between those two points in time, it shows three 5*s before and then three 1*s after that snapshot. If it had shown an 11/7 split of stars that could still be three 5*s and three 1*s, but it could also be two 4*s and a 3* and then a 4*, a 2*, and a 1*. Or a bunch of other extremely unlikely splits. But what it actually showed was that those 18 stars among 6 votes came in the exactly predicted three 5*s and three 1*s.

For it to be anything else, the 4.91 (103 Votes) would have to be a complete algorithmic hallucination that corresponds to no vote totals that have ever existed. And even if it is an algorithmic hallucination, the three 5*s and three 1*s split is still by far the most likely. Because 5* ratings are the most common in general and 1* ratings are the most common troll vote and there is no evidence at all that any other kind of vote was cast.

If you didn't know anything about what Literotica voting patterns looked like, and you were told that six votes were cast and the total was 18 stars, you'd probably assume it was a nice like Gaussian curve at 5, 4, 3, 3, 2, 1. But that's INCREDIBLY unlikely given what we know about Literotica voting and trolling patterns. 3* ratings are the least used ratings on this site.

So you think that the other one has picked up 34 votes in the time between the two toplists updating? Again, you're latching onto whatever you think can confirm your hypothesis, spitting out numbers, and ignoring anything that doesn't fit your premise.

That story did not pick up 34 votes and go up .02 in whatever imaginary brief window there was between the two toplists updating. The 12 month list is simply wrong. K.I.S.S.

( For the record, I'm not attempting to convince RG of anything. That's a lost cause. All I'm doing is dropping reality checks so nobody else gets confused by the teletype regurgitation of numbers and gets pulled into this conspiracy theory, setting off yet another social contagion. )

 

[SimonDoom]

If this thread was a dog, a vetenarian would have put it out of its misery by now.

A dog chasing its tail, not understanding why others don't sympathize with its obsessive need to catch it.

 

[lovecraft68]

If this thread was a dog, a vetenarian would have put it out of its misery by now.

I keep seeing it as more of a hamster in a wheel visual.

 

[Dodola]

Obsession abounds.

Neither side will let the other have the

Last

Word.

 

[RocketGrunt]

So you think that the other one has picked up 34 votes in the time between the two toplists updating? Again, you're latching onto whatever you think can confirm your hypothesis, spitting out numbers, and ignoring anything that doesn't fit your premise.

That story did not pick up 34 votes and go up .02 in whatever imaginary brief window there was between the two toplists updating. The 12 month list is simply wrong. K.I.S.S.

( For the record, I'm not attempting to convince RG of anything. That's a lost cause. All I'm doing is dropping reality checks so nobody else gets confused by the teletype regurgitation of numbers and gets pulled into this conspiracy theory, setting off yet another social contagion. )

I don't know anything about the other one's extra 34 votes. It may well have reached a 240 vote count at some point in the last twelve months. Maybe there were a bunch of votes that got swept. We haven't been tracking that one at all. So whether it refers to a specific point in time or not is something neither of us are qualified to have an opinion about.

The One Whore's Town Chapter 3 story definitely did pass 103 votes this week. It is definitely a real vote total that it had, and definitely a vote total that it had between when I saw it at 100 votes and when you saw it at 106 votes. It is metaphysically impossible to add votes to go from 100 to 106 without passing through 103 along the way. So why would we think that the 103 vote report isn't real? It's a vote total that we have absolute certainty really happened and we have a very good idea of about when it happened.

I just don't understand how "the progression of numbers that is most likely and is also algebraically derivable from the available checked vote totals didn't happen" could possibly be the KISS explanation. You're the one spinning weird conspiracy theories. You asked us to look at a specific group of stories and watch how the numbers played out. We did that. We have enough data points that we can derive the exact ratings given and the exact order that they were given, and it matches the prediction of scripted trolling exactly. Like, EXACTLY exactly.

 

[SimonDoom]

That one vote. That one vote. The one that knocked my story from 4.92 to 4.91. I must figure it out. I must expose the nefarious plot behind it. I must expose the awfulness and injustice of a site that allows a scoring system that allows this to happen to me. I must not stop in my investigation of how it happens. Justice depends upon my investigation. No one can stop me. Never mind that my scores are higher than most of the other authors to whom I am complaining. Never mind that no one cares or that this injustice has no impact on the effectiveness of the site to deliver a product that its readers want. It's the injustice that matters. It gnaws at me. Listen to my story of woe.

Yes, barkeep, I'll have another. Could you make it a double?

 

[RejectReality]

I don't know anything about the other one's extra 34 votes. It may well have reached a 240 vote count at some point in the last twelve months. Maybe there were a bunch of votes that got swept. We haven't been tracking that one at all. So whether it refers to a specific point in time or not is something neither of us are qualified to have an opinion about.

The One Whore's Town Chapter 3 story definitely did pass 103 votes this week. It is definitely a real vote total that it had, and definitely a vote total that it had between when I saw it at 100 votes and when you saw it at 106 votes. It is metaphysically impossible to add votes to go from 100 to 106 without passing through 103 along the way. So why would we think that the 103 vote report isn't real? It's a vote total that we have absolute certainty really happened and we have a very good idea of about when it happened.

I just don't understand how "the progression of numbers that is most likely and is also algebraically derivable from the available checked vote totals didn't happen" could possibly be the KISS explanation. You're the one spinning weird conspiracy theories. You asked us to look at a specific group of stories and watch how the numbers played out. We did that. We have enough data points that we can derive the exact ratings given and the exact order that they were given, and it matches the prediction of scripted trolling exactly. Like, EXACTLY exactly.

No, it doesn't. It requires 3 magical 5s that defy anything resembling common sense to work. All my hypothesis requires is for people to be assholes.

And, poor Mage of Dockside has finally fallen to the troll army. At least I've seen the toplists update more than once in a 24 hour period now, so I know that's a thing, where it historically never was.

 

[FrancesScott]

The post-competition-winning vote suppression of my latest story (currently on 4.81 after 98 votes) has at least spared me the post-top-lists vote suppression that it would have incurred had it retained its previous 4.87 rating through 100 votes. I am grateful for small mercies!

 

[SmilingLez]

So I decided to look at the top lists. Don't know if it's broken, but definitely skewed. It seems like ten 5 stars from 10 readers are rated higher than 4.75 stars from 100 readers with 30 readers giving 5 stars

 

[RocketGrunt]

No, it doesn't. It requires 3 magical 5s that defy anything resembling common sense to work.

There is literally no reason to doubt the existence of the three 5* votes. Rather, there were a total of 8 votes. I saw a vote increment that proves a 5* rating was given. We all agree that one doesn't defy anything related to common sense. You saw a one vote increment that proves a 1* rating was given. That one is obviously "inauthentic" but we all agree it happened.

In between, there were six more votes, and we know they add up to 18. The simplest explanation is that three match the first vote and three match the 8th vote. It genuinely doesn't defy any common sense for the votes that come between the votes we have one increment series on to match those two votes. If they were any other numbers, we'd need more stories to explain the ones in the middle.

And then we have an extra data point that very much looks like a snapshot of the vote tally exactly in the middle of those six. And if it is correct, it would absolutely prove that the first four were the same and the second four were the same. So the thing that was obviously true from the beginning is simply still obviously true.

I genuinely do not know why you think it defies all common sense, or any common sense, for the numbers we weren't sure about to be exactly the same as the numbers we were sure about.

So I decided to look at the top lists. Don't know if it's broken, but definitely skewed. It seems like ten 5 stars from 10 readers are rated higher than 4.75 stars from 100 readers with 30 readers giving 5 stars

The old toplists are a 250 story list that displays stories that have been rated at least 100 times in order from highest to lowest and in the case of ties gives an ordering by total number of ratings.
The new layout has a "Popular Stories List" that also orders things by average rating and then by number of ratings for ties, but has a much lower threshold of required total votes. If you click on "more popular stories" it sends you to the old toplist page because the new page isn't live and this website is full of quarter century old spaghetti code.

If you look at most categories, you will find MUCH higher ratings on the Popular Stories List than on the old toplists. That is because there are trolls that procedurally dump 1* ratings on stories that enter the toplist of any category with a rating above 4.85. This means that the Popular List will show ten stories that have less than 100 ratings and averages around 4.9-5.0 instead of 4.84. On the Loving Wives, Taboo/Incest, and Mature boards specifically, those same troll procedures attack stories above 4.84 that have less than 100 ratings and are not on the toplist, so the Popular List and the Toplist are the same.

 

[Voboy]

That one vote. That one vote. The one that knocked my story from 4.92 to 4.91. I must figure it out. I must expose the nefarious plot behind it. I must expose the awfulness and injustice of a site that allows a scoring system that allows this to happen to me. I must not stop in my investigation of how it happens. Justice depends upon my investigation. No one can stop me. Never mind that my scores are higher than most of the other authors to whom I am complaining. Never mind that no one cares or that this injustice has no impact on the effectiveness of the site to deliver a product that its readers want. It's the injustice that matters. It gnaws at me. Listen to my story of woe.

Yes, barkeep, I'll have another. Could you make it a double?

Plot bunny: hatched.

A nice noir for the next Crime And Punishment event.