I see my score is 4.18

[AwkwardlySet]

So this is my idea. But first of all, I hope we understand this is all just wishful thinking and discussing of some interesting ideas that will surely never happen.

The basic idea is quite simple. Readers would be able to vote by clicking a simple thumbs-up button at the end of the story. There would be no thumbs-down option, just the option to "like" the story, or not to vote at all. Then, the score would be derived by dividing the number of thumbs-ups by the total number of views, say something like 100/1000. This "score" would then be shifted into values that are easier for readers to visually track. For example, the above 0.1 --> 10, etc. Colors could also be used to display the value of the score.

Now, to explain and expand on this idea. There are a couple of reasons why fraudulent voting is so powerful in the present system.

1. The scores are so ridiculously shifted towards 5 that every 1* has much more power than it should.

2. In the present system, where unregistered users are allowed to vote, there is no way to tie the vote to the reader who cast it, so by using simple technical tricks, readers can cast as many votes as they like on each story.

These two reasons make 1-bombing tremendously effective, and while 5-bombing is much less effective since #1 doesn't apply, a patient and committed person could still efficiently 5-bomb a story given enough time.

The system I presented above doesn't have any of these problems, but it requires only registered users to vote, and only their views to count towards determining the score. The story could still display, as additional information, the total amount of views, both from registered and unregistered users, but only the views of the registered users would count towards the score. This is to disallow unregistered users from spamming "refresh page" on any story just to boost the view count and thus fraudulently lower the score.

Now, how is my system more fair and accurate than the present one?

First, there is no way to downvote. All anyone can do is create a "view" without "liking" the story. And that's it. The effect a single person can inflict on each story is minimal (I won't get into exact math right now). Each user can "view" and "like" a story only once. No one can 1-bomb or 5-bomb a story, especially considering that the registration process could be made to take a day or two. No one is going to go through all the fuss of creating an email and registering just so they could add one "view" or one "view" + one "like."

Finally, how would these scores be ranked? There should be a minimal number of "likes" for the score to count, similar to the way SOL works, and even the red H on Lit. The percentile-based ranking system seems natural with this idea, but other approaches can work too.

I'll stop here. There is math I avoided going into to keep this simple and as short as possible, but it can be easily shown that this way of deriving scores would be much more resilient to any fraudulent actions. Readers would still be the judges of which stories are "the best," but no individual would be able to single-handedly influence the score in any notable way.
So all the butthurt readers and forum trolls, but also devoted fanboys and girls, wouldn't be able to snipe stories in the top lists anymore, and wouldn't be able to sway the contest scores.

 

[SimonDoom]

Can you think of any examples of this method having been tried before? I don't see why it would yield a "score" that would convey the information people would want it to, but if it's been demonstrated to work I'd be willing to consider the evidence.

 

[NotWise]

@AwkwardlySet, it seems like viewers who move on without clicking the "thumbs up" lower the score, whether they read the story or not, and whether they like the story or not.

How does having an account make a reader more qualified to express an opinion?

 

[pink_silk_glove]

You are contradicting yourself here. Yes, some will care, and that means yes, they will be more likely to read top-rated stories, just as they are now. That's a good thing, not a bad thing. No author can fairly claim: "Now that the story rating system is more accurate my story is less likely to be read than before."

Actually there will be some people very upset. Those who are stroking themselves and walking around here with their noses in the air because their 4.6s give them Red Hs and they will lose most of those, even though that can easily be replaced with something more accurate like a tiered Hot (70th%) Hott! (80th%) and HAWT!! (90th%) which they still have a chance to make 3rd tier. These folks don't want accuracy. They just want their average scoring story to be stamped Hot so that they can continue believing that they are the reincarnation of John Lange or Anais Nin.

 

[AwkwardlySet]

How does having an account make a reader more qualified to express an opinion?

It doesn't. The requirement exists to keep the votes authentic and to combat fraudulent voting.

 

[AwkwardlySet]

I don't see why it would yield a "score" that would convey the information people would want it to, but if it's been demonstrated to work I'd be willing to consider the evidence.

It practically eliminates fraudulent voting. Also, unlike SOL, where a 1* vote from a registered user still has a significant impact, each thumbs-up vote, or the lack of one, has a balanced impact on the score.
I've no idea if this system is used elsewhere. I came up with this myself. But I believe the math is in my favor.

 

[old_prof]

Three points:

I don't think it is so much easier to change IP's than to create multiple accounts that this will reduce 1-bombs notably. But it will lose the voice and reaction of casual readers.

This is so sensitive to exactly how views is captured and how people choose to read. I have no idea if a two (Lit) page story gets two views or one. How about if someone has it in a tab and the browser chooses to reload that page? What is someone starts reading a story, puts is away and reloads the page later to finish? Do we have a great sense that page views is any more accurate of a measure than anything else is?

Yours is more a measure of a form of engagement than anything else. Imagine a stroker story, Maybe it works really well and the average reader is satisfied, physically at least, before reaching the voting mechanism. That story is rated no better and maybe worse than one that fails to achieve its outcome for the reader more effectively. People's decision to vote or not vote is a complicated decision, generally unrelated to any particular quality of the story (except maybe for certain strokers).

 

[AwkwardlySet]

Three points:

I don't think it is so much easier to change IP's than to create multiple accounts that this will reduce 1-bombs notably. But it will lose the voice and reaction of casual readers.

No offense, but this makes no sense. It takes a few clicks to change an IP address, yet it takes creating a new email on a server and filling in all the data, then registering on Literotica and verifying the email, and then waiting for a day or two for the approval. The site could even create a much simpler version of the swiping algorithm compared to the one it uses now to occasionally delete the alt accounts by checking whether the IPs on any two accounts closely match. So, a fraudulent voter would need to both create alt accounts and shift to a VPN every time they use the alt account. No one is that insane.

This is so sensitive to exactly how views is captured and how people choose to read. I have no idea if a two (Lit) page story gets two views or one. How about if someone has it in a tab and the browser chooses to reload that page? What is someone starts reading a story, puts is away and reloads the page later to finish? Do we have a great sense that page views is any more accurate of a measure than anything else is?

Story views could be recorded the same way as here, by clicking on the story and loading the first page. But the "view" would be linked to a registered user, and it would count only once, so clicking reload on a story wouldn't have any effect. The views of unregistered users, as I said, wouldn't count towards the thumbs-up/views score, so if they clicked reload a thousand times, it still wouldn't matter one bit.

 

[pink_silk_glove]

The basic idea is quite simple. Readers would be able to vote by clicking a simple thumbs-up button at the end of the story. There would be no thumbs-down option, just the option to "like" the story, or not to vote at all. Then, the score would be derived by dividing the number of thumbs-ups by the total number of views, say something like 100/1000. This "score" would then be shifted into values that are easier for readers to visually track. For example, the above 0.1 --> 10, etc. Colors could also be used to display the value of the score.

So if someone clicks the story 3 times (to read it again every few months or to read a longer story in 3 sittings possibly from different devices) and they give it a 'like' they actually drive the rating down? Or if a writer clicks on one of his own stories every few months he drives his own rating down? Can't say that I'm into that at all.

I do like the idea of thumbs up thumbs down better than the 1-5 scale with absolutely no rubric whatsoever. I'm on record here stating that before.

On the whole, your system kinda reeks of 'no negative feedback allowed' which is silly. We're adults here, this isn't kindergarten.

2. In the present system, where unregistered users are allowed to vote, there is no way to tie the vote to the reader who cast it, so by using simple technical tricks, readers can cast as many votes as they like on each story.

Banning unregistered voting is off the table. It won't do anything to significantly curb fraudulent voting. The vast majority of bombs are one person with one or two accounts running down a hated author's list. I know this because that's how y'all (trolls in AH) do it to me. My votes rarely move so I can see exactly when you do it. I can even tell who does it by how many of my newest only stories get hit (how long since the last time that that particular hater bombed me). Just stop this silly crusade. Banning any kind of unregistered feedback (votes are feedback) does way more harm than good.

First, there is no way to downvote.

That's the issue. Knowing your personal crusade for months and months, and seeing your outlined system, it's quite obvious that this is simply a self-serving way to protect your own scores and keep your own ego happy. The whole thing reeks of, "I'm going to publish my story and bask in the sunshine of everyone telling me how good it is." This system won't improve the accuracy of votes, in fact in might reduce it since it will cut the volume in half which will increase the variance (create more false outliers). This system is also designed for writers to get a score instead of helping readers choose stories and this counter productive premise is the entire foundation of your system.

Sorry, don't like it. (Thumbs down. : P )

 

[pink_silk_glove]

But the "view" would be linked to a registered user,

Now that would increase server space. Imagine the text string in an account that hit 50 stories per week? It would be 100 times longer than the one that tracks the account's story votes (as votes to views is roughly 1:100).

 

[NotWise]

It doesn't. The requirement exists to keep the votes authentic and to combat fraudulent voting.

Do you believe anonymous voters (who may be registered but not logged in) are inauthentic frauds? Does logging in alter someone's character?

 

[AwkwardlySet]

Do you believe anonymous voters (who may be registered but not logged in) are inauthentic frauds? Does logging in alter someone's character?

I've no desire to discuss human nature. All I am saying is that in the system I described above, they wouldn't be able to act in a fraudulent way because of the system design. Unregistered users would still be able to read the stories and comment on them, just not to vote. The site could even encourage users to register (and thus increase its own value) by giving them additional benefits, such as we discussed here in many previous threads - voting, more interaction with authors, robust control panel and interface, etc.

 

[old_prof]

No offense, but this makes no sense. It takes a few clicks to change an IP address, yet it takes creating a new email on a server and filling in all the data, then registering on Literotica and verifying the email, and then waiting for a day or two for the approval. The site could even create a much simpler version of the swiping algorithm compared to the one it uses now to occasionally delete the alt accounts by checking whether the IPs on any two accounts closely match. So, a fraudulent voter would need to both create alt accounts and shift to a VPN every time they use the alt account. No one is that insane.

You are just not being creative enough to think how to beat this because you are so convinced it is unbeatable. It is trivial to create new accounts for anything. There are plenty of tools that will do it for you. And you can use an account repeatedly until it gets detected, rather than having to change IP for each vote for a story. It might be trivially more challenging but if people are so determined to undermine your vote, they will find a way.

 

[AwkwardlySet]

Now that would increase server space. Imagine the text string in an account that hit 50 stories per week? It would be 100 times longer than the one that tracks the account's story votes (as votes to views is roughly 1:100).

It's still all text, and server space is the cheapest thing about any website.

Let's say that Literotica has ten million registered users, and let's also assume that a user has clicked on about 100,000 stories on average, which is probably too much. Say that all the stories are tracked by their number, and say that the number is 20 bits long (let's round it up to 24 for ease of calculation, which is sufficient for a bit over 16 million stories) By the way, Lit has managed to amass somewhere around 600k+ stories so far.
So, ten million users x 100,000 stories x 24 bits that are needed for each story to be identified with a different number. The additional storage space needed is around 3TB, which is one SSD practically. For a website with 10 million registered users, the additional cost is negligible.

 

[pink_silk_glove]

I've no desire to discuss human nature. All I am saying is that in the system I described above, they wouldn't be able to act in a fraudulent way because of the system design.

First, your entire premise, system and crusade is all about human nature. You want to rid the world of fraudulent votes, but fraudulent voting is all about human nature.

Second, for the umpteenth time, requiring all voting to be from registered accounts will do very little to curb fraudulent voting yet will cause undue harm to legitimate feedback.

 

[pink_silk_glove]

It's still all text, and server space is the cheapest thing about any website.

Let's say that Literotica has ten million registered users, and let's also assume that a user has clicked on about 100,000 stories on average, which is probably too much. Say that all the stories are tracked by their number, and say that the number is 20 bits long (let's round it up to 24 for ease of calculation, which is sufficient for a bit over 16 million stories) By the way, Lit has managed to amass somewhere around 600k+ stories so far.
So, ten million users x 100,000 stories x 24 bits that are needed for each story to be identified with a different number. The additional storage space needed is around 3TB, which is one SSD practically. For a website with 10 million registered users, the additional cost is negligible.

Still, that much memory just to keep track of your personal self-serving vanity project.

 

[NotWise]

I've no desire to discuss human nature. All I am saying is that in the system I described above, they wouldn't be able to act in a fraudulent way because of the system design. Unregistered users would still be able to read the stories and comment on them, just not to vote. The site could even encourage users to register (and thus increase its own value) by giving them additional benefits, such as we discussed here in many previous threads - voting, more interaction with authors, robust control panel and interface, etc.

Requiring readers to be logged in to participate in voting would probably make a big hit on the number of votes cast. I say "probably" because we don't know now whether or not a voter is logged in. I suspect that a minority of the people viewing and potentially voting are actually logged in at the time. Reader participation would drop. Feedback would drop.

The bigger problem is your use of views. Views are a measure of traffic, and that's all. They aren't reads. Any time a viewer leaves without voting it would work as a down vote. Your system makes it extremely easy to down vote. They don't even have to think about it. They might even love your story and still register a down vote. If someone did want to attack a story, all they'd have to do is register a view over and over.

I've monitored votes on my stories for years now, and I don't think that fraud or invalid votes are the problem you think they are. And down votes aren't a problem. They're an opinion. In your system, down votes would be constant and unavoidable, and not even intentional.

 

[lovecraft68]

I've no desire to discuss human nature.

Human nature has driven this entire thread.

 

[ElectricBlue]

I've monitored votes on my stories for years now, and I don't think that fraud or invalid votes are the problem you think they are. And down votes aren't a problem. They're an opinion. In your system, down votes would be constant and unavoidable, and not even intentional.

Agree. I've kept an eye on my scores for eleven years now (seeing trends over time, but not "monitoring" as such, with detailed records), and occasionally I see a blip in numbers, but nothing that ever suggests a concerted campaign against either the stories or me as an author.

I nearly always see the clown votes straight out of the gate (as we all do), but I also see the score adjusted every time a sweep goes through. Vote counts drop, the score always increases. It's part of the furniture, it happens to all of us.

And that's the point, it happens to all of us. It's a universal element of the system, we're all affected equally, so there is no advantage or disadvantage. No-one is "suffering" when compared to the next guy in terms of story exposure.

The same random bloc of readers that bothers to give a story a score (1 vote per 100 views) is the proxy input for the tens of thousands of readers who read stories, but don't bother giving a score. And here we have a proposal to cut that scoring bloc in half or worse, by preventing anon voting? That's nuts.

 

[AwkwardlySet]

Human nature has driven this entire thread.

and it was also the cause of the issues that gave birth to it.
My whole point in this thread was that we can't hope to change human nature. Some people will always try to hurt others, to game the system, to exploit something for their own benefit. So I tried to focus on the only thing that made sense - discussing the system that would place some technical limitations on those who want to exploit and cheat. And even that only for the sake of nerd-talk. I am fairly sure everyone here understands that nothing we discuss will have any impact on the way the website works.

 

[old_prof]

But all of your "proof" assumes users will perform as rational actors according to your internal blueprint of what rational actor is. And you reject any consideration that this is not the way they will work.

This ignores the 800 lb gorilla in the room,. The site's goal is to maximize page views for total uses, not just registered users. What you are suggesting may slightly increase the number of user's registering (maybe much more in spurious registrations), but will almost certainly drive down non-registered usage. The voting is a way to make non-registered readers feel engaged, keeping them tied to the platform better.

Okay, this is the last time I will be sucked into this.

 

[SimonDoom]

It practically eliminates fraudulent voting. Also, unlike SOL, where a 1* vote from a registered user still has a significant impact, each thumbs-up vote, or the lack of one, has a balanced impact on the score.
I've no idea if this system is used elsewhere. I came up with this myself. But I believe the math is in my favor.

I don't think it's a matter of math; it's a matter of predicting voter behavior and what the resulting score will mean. I have no confidence that it will yield a result readers will want. I have no clue how readers will respond to it by voting. I don't think anyone does without evidence of past use.

Here are my objections:

1. I don't want to vote just thumbs up. I want to be able to vote 1-5. I think most voters would prefer the current option to yours. Maybe I'm wrong, but if I'm right, and I think I am, that torpedoes the idea right there. It makes no sense to create a system people don't actually want.

2. I question how significant the fraudulent voting problem is. It happens, but how much? I've been here 8 years and I'm not confident it's ever happened to me. I've been downvoted--I'm sure of that--but I don't see that as inherently fraudulent. It's not enough to say that the system has the weakness of creating the risk of fraudulent voting, and your system eliminates the risk. There's no reason to replace the system unless we have good reason to believe that it actually happens enough that it's a big problem. I don't see it.

3. I have no idea what the resulting scale would look like, and I'm not sure that a system based on thumbs-up:view ratios would yield the information I would want as a reader. Maybe, but maybe not. Why should I believe that it would if it hasn't been tried and shown to work?

4. You still have the downside of cutting half of current voters, or more, out of voting. That's a big disadvantage and you'd better have something that's CLEARLY better before replacing it. It seems purely conjectural to me that this system would be better, and for me personally it probably would not be, so I don't see the threshold for change being met.

 

[AwkwardlySet]

This ignores the 800 lb gorilla in the room,. The site's goal is to maximize page views for total uses, not just registered users. What you are suggesting may slightly increase the number of user's registering (maybe much more in spurious registrations), but will almost certainly drive down non-registered usage. The voting is a way to make non-registered readers feel engaged, keeping them tied to the platform better.

You are working under the wrong assumption here. I'm not advocating that the system I presented would be better for Literotica's profits and traffic. I have merely presented a system that would, in my opinion, produce more fair and accurate scores, and nothing more than that. What its impact on the traffic and number of visitors would be, we can only speculate. No one knows for sure, of course.

This is just a math model for a more accurate scoring system. I never truly tackled its impact on the traffic, long-term viability, etc. That would require too much guessing and speculation.

 

[AwkwardlySet]

I don't think it's a matter of math; it's a matter of predicting voter behavior and what the resulting score will mean. I have no confidence that it will yield a result readers will want. I have no clue how readers will respond to it by voting. I don't think anyone does without evidence of past use.

Here are my objections:

1. I don't want to vote just thumbs up. I want to be able to vote 1-5. I think most voters would prefer the current option to yours. Maybe I'm wrong, but if I'm right, and I think I am, that torpedoes the idea right there. It makes no sense to create a system people don't actually want.

2. I question how significant the fraudulent voting problem is. It happens, but how much? I've been here 8 years and I'm not confident it's ever happened to me. I've been downvoted--I'm sure of that--but I don't see that as inherently fraudulent. It's not enough to say that the system has the weakness of creating the risk of fraudulent voting, and your system eliminates the risk. There's no reason to replace the system unless we have good reason to believe that it actually happens enough that it's a big problem. I don't see it.

3. I have no idea what the resulting scale would look like, and I'm not sure that a system based on thumbs-up:view ratios would yield the information I would want as a reader. Maybe, but maybe not. Why should I believe that it would if it hasn't been tried and shown to work?

4. You still have the downside of cutting half of current voters, or more, out of voting. That's a big disadvantage and you'd better have something that's CLEARLY better before replacing it. It seems purely conjectural to me that this system would be better, and for me personally it probably would not be, so I don't see the threshold for change being met.

Once again, I have no idea why everyone assumes that I am trying to present a Holy Grail for Literotica here. I am merely presenting a math model that would lead to more accurate scores, not because it's ingenious, but simply because it allows far less room for fraud. That's all there is to it.

I am certainly not discussing whether or not readers would be happy to switch to a system that gives them fewer options. Logic says they would probably dislike it, at least initially. No one likes having fewer choices, especially after having more for such a long time. Would that drive readers away? Would they adapt? Would they eventually warm up to the system? I have no idea, and that's why I am not talking about those aspects at all.

 

[pink_silk_glove]

You are working under the wrong assumption here. I'm not advocating that the system I presented would be better for Literotica's profits and traffic. I have merely presented a system that would, in my opinion, produce more fair and accurate scores, and nothing more than that. What its impact on the traffic and number of visitors would be, we can only speculate. No one knows for sure, of course.

This is just a math model for a more accurate scoring system. I never truly tackled its impact on the traffic, long-term viability, etc. That would require too much guessing and speculation.

If it does not allow a negative vote, it's NOT more accurate than what we have.

This is all about you not wanting anyone to give you a negative vote. This is all about you protecting your own ego.