Inner workings of voting

[palisa]

This may be the wrong forum for this question, but the old hands seem to hang out here.

My maiden effort of a story has been just sort of sitting there, with perhaps a dozen or two new reads per week, no public comments since the first few days, only one anonymous comment. It got 30 votes, but no additional votes in the past two weeks or so.

I checked again today, and saw that three votes had been removed and the score was up a bit. Is this common? Anybody care to explain what happened (or hazard a guess)?

Thanks.

 

[emap]

Most only vote for new stories, honestly you should be happy you got 30 votes, or well less I guess now. I have a number of my stories with less than 20 votes.

As for comments, honestly the public comments generally don't happen unless say you ask for them at the beginning of the story. Even then don't expect to get any past the first week.

Private comments follow the same pattern, I rarely get more than 5, usually less.

Oh and the lost votes are repeat votes from the same computer, they scan or whatever every once in a while and you will notice votes going away, I beleive it is once a month.

 

[snooper]

... the lost votes are repeat votes from the same computer, they scan or whatever every once in a while and you will notice votes going away, ...

As I understand it, an attempt is made to delete troll votes. It is a difficult problem because people (like me) who are using dial-up have a different IP address (within a range available to the ISP concerned) every time we log on. Troll votes are defined as a series of very low votes from user(s) who have not actually read the story.

What I fail to understand is why anyone bothers even to look at the voting on a story. I cannot know what criteria are applied by each voter, and without some way of calibrating each voter I don't know how any meaningful value can be put on any one vote, let alone an aggregate of 30 votes added together. To take a simple example, if I have an average of 3 on two votes is that a 1 and a 5 (somebody hated it and somebody loved it) or a 3 and a 3 (two people both thought it was mediocre).

This topic is (or at any rate, used to be) discussed at great length and quite often on the Authors' Hangout.

 

[emap]

Well as near as I can tell the only person who can see the vote tally is the author until the story gets into the top 500 of that section at which point I do beleive it shows the average.

Which more or less means the only people who give a flying leap about them are the authors. I don't know why I always base the quality on the feedback I receive, simply because as you said the tally doesn't really tell you anything.

 

[JPMMURPHY]

All true. One thing I might add from personal experiece - the key to continued voting and commentary is .... writing. Each time you put a new story up, if people enjoy it, many follow the link to your profile to see what else you've written. They might click on something, vote, comment, and check your profile again. Every time you appear in the New Story list is like a billboard saying - here I am!

My suggestion - keep writing. If you find your audiance (we all share the same audiance pool - but different segments) they will be loyal and prolific in their votes and comments.

BTW - I didn't check out your story but 30 votes on a first effort is nothing to sneeze at. Enjoy it.

 

[palisa]

Thanks for the comments. The point of my post was the question about the mechanics of the changing vote total; I included the comments about the reads and votes as background. It seemed a bit odd to me that after the story had lapsed into obscurity there would suddenly be that adjustment downward in the vote count. Since the average went up, it was clear that some low votes had been deleted, but I didn't know why.

The site rules say the author is responsible for deleting inappropriate public comments. Again, not knowing whether such comments are common or whether an offensive comment would indeed put the author in the soup, I've been checking every few days.

A couple of authors I've edited have asked me about such site-specific matters, and I have to reply that I don't know much about how this site works.

I agree that it would be nice to know the spread of votes rather than just the average. But I also wonder a bit about the disdain, feigned or otherwise, for tracking the reception of stories. Granted, the system isn't anywhere near perfect, but there's no money involved, which would be the best indicator of how a story has been received. If you're not looking for reader acceptance of a story, why write it? I suppose some people prize self-expression above reader reaction, but I've got my doubts about how useful or satisfying that would be.

Almost all ISPs, not just dial-up systems, use dynamic IP addresses. I don't quite follow the logic of how that makes "troll votes" more susceptible, not less, to deletion, but I'll search on the term to find past threads on the subject.

Later: OK, after having read the other three threads that searching on "troll vote" brings up, I'm still not sure what the term means. The consensus seems to be any vote of 3 on down. While that's obviously correct when applied to MY stories, I'm not so sure it's applicable to YOURS. (Is the smiley enough to establish tongue-in-cheekness?)

 

[MarshAlien]

I don't believe that I am breaking any confidences by sharing this PM that I received from Laurel last night. I had written to ask whether it was possible that a story of mine, which was posted in February and had survived numerous sweeps between now and then (they usually take place during contests such as the Halloween Contest going on right now), had lost 240 of its 520 votes during the last 24 hours. She explained:

We have a script that tags fraudulent votes (fraud votes being anytime someone tries to vote for a story more than once by trying to get around our system). If you lost votes and your score went up, chances are that fraudulently placed lower votes were removed. If your score went down, it's possible that overzealous fans were attempting to "help" you by submitting multiple up-votes on your stories.

Hope this helps.

 

[emap]

Troll votes as near as I can figure it are people who go in and give you a vote of 1 and/or berate your story for no particular reason other than they disagree with the subject matter.

Of course us authors being human, many I would assume figure any score below a 4 is a troll vote.

Personally I figure any time I get comments, public or private I did a good story. Though I suppose I am extra perverted and probably should be shut up in one of those mental health homes. Well personally I think a full on insane asylum would be better, though sadly there aren't any left.

 

[snooper]

Well as near as I can tell the only person who can see the vote tally is the author ...

Everyone can see the vote tally on every story.

If you go to http://www.literotica.com/stories/index.php and look in the category of the story you want, there is a list of all the stories in that category in alphabetical order of title. The very first entry in Erotic Couplings is:

"F" Is For Female - A man's response to an unusual request.
Submitted by Phrenetic_Ice (4.10) 11/11/04

The number in parentheses is the current vote for that story.

... Almost all ISPs, not just dial-up systems, use dynamic IP addresses. I don't quite follow the logic of how that makes "troll votes" more susceptible, not less, to deletion, ...

I meant that changing IP address and voting without logging on makes them less susceptible to detection.

 

[palisa]

I meant that changing IP address and voting without logging on makes them less susceptible to detection.

Yes, you did. I misread what you posted.

I just noticed that a fourth vote was deleted today and the score is up again. Climbing the charts through the power of subtraction.

 

[Jenny_Jackson]

The mysteries of the vote-score has baffled Lit writers for generations. I've come to the concludsion that the scores are a thing of witchcraft, Dungeon Siege, Diablo and Demonism.

I could write a story about it, I think.

 

[desertkat]

Public and private comments

I like the voting and I think they should leave all votes except the duplicate ones alone. A bad vote simple means I need to improve something. If someone disagrees with my subject matter, it is their right. But I do believe that if you have chosen to read a BDSM story and the beating of a sex partner who is tied up offends you.... WTF were you doing reading a BDSM story?! Same goes for non consent/reluctance, if rape isn't in your fantasy realm DON'T READ IT! And for heavens sake don't berate the author for writing it. We each have different things that trip our triggers, what makes you the judge and jury? I receive lots and I mean lots of private comments even on stories I have posted almost 2 years ago. In fact I just got another e-mail on a story I did last year. I like all comments - even the bad ones they help to guide my growth. And I want to thank all of the readers who have taken the time to send me e-mails and post comments. It is because of the almost overwhelming response to these few stories I have posted here that I have actually had the courage to pursue a writing career. I just signed my first contract a few weeks ago. I personally am very greatful for this site. It allows me study personalities, techniques, plots (or lack thereof) and to get unbiased opinions of my work. Friends and family tend to sweeten their comments, an annonamous reader tells it like they feel it.

 

[CorsetLvr]

And then on the other hand...

Take a look at the number of votes vs. the number of views on your story. If your story is anything like mine, the percentage of readers that take the time to vote is significantly less then 1%. Is a sample size of less then 1% statistically significant? Psychologists will tell you that the majority of voters are those that feel strongly about a story, wither positively or negatively. The rest are within the "slient majority."

As a writer I have anguished over my score in the past. I have gotten past that. My ego does make me keep an eye on the score, but does what less then 1% of myreaders think about my stories keep me up at night? Obvioulsy no.

 

[sr71plt]

And then on the other hand...

Take a look at the number of votes vs. the number of views on your story. If your story is anything like mine, the percentage of readers that take the time to vote is significantly less then 1%. Is a sample size of less then 1% statistically significant? Psychologists will tell you that the majority of voters are those that feel strongly about a story, wither positively or negatively. The rest are within the "slient majority."

As a writer I have anguished over my score in the past. I have gotten past that. My ego does make me keep an eye on the score, but does what less then 1% of myreaders think about my stories keep me up at night? Obvioulsy no.

I think a good number of readers who vote don't do so even because they had a strong reaction to the story itself. I think some readers vote on most stories they read and most don't--just a habit of clicking the vote numbers or not--across the board, not as a particular reaction to a particular story.

 

[LadyCibelle]

I think a good number of readers who vote don't do so even because they had a strong reaction to the story itself. I think some readers vote on most stories they read and most don't--just a habit of clicking the vote numbers or not--across the board, not as a particular reaction to a particular story.

I have to agree with Sr on this one. Some readers are just more encline to vote, they'll vote on everything whether they like it or not; it's in their make-up I guess. But, I also agree with CorsetLvr that a big majority of those voters will be the ones who are deeply affected by your story, be it good or bad, if you make them react strongly (no matter the reaction), they'll vote accordingly to their reaction.

 

[snooper]

... Is a sample size of less then 1% statistically significant? ...

That depends on the sample selection process. Political opinion polls with carefully selected stratified random samples of the population go as low as one in fifty thousand of the voters (0.002%) and are accurate to quite close limits (typically within 3% of the actual voting).

The reason a 1% sample of Lit readers is not significant is that it is self-selecting. Only people with strong views will vote (apart from the relatively small number of habitual voters); this means that the extremes are over-represented, especially those with strong negative views.

In a crude voting system (like politics or Lit) it is difficult to work out why people are voting as they do. For example if Lit introduced compulsory voting then there would be an important group voting 1 on every story they read as a protest against compulsory voting, and another group voting 5 on every story for the same reason!

 

[sr71plt]

this means that the extremes are over-represented, especially those with strong negative views.

Agreed with you except for this statement. I think it's obvious that voters with strong positive views are on the "overrepresented" side on Lit. (which is nice--I wouldn't knock that). The overall voting average is well over the "4" mark, when, if negativity prevailed in voting, the voting average would be no higher than the "2" mark.

 

[fieryjen]

That doesn't necessarily mean that though. You assume that there are just as many strong positive as strong negative views. However, if there are many more strong positive views, the strong negative views can be way overrepresented and still end up not influencing a vote as much as the positive views.

 

[CorsetLvr]

That depends on the sample selection process. Political opinion polls with carefully selected stratified random samples of the population go as low as one in fifty thousand of the voters (0.002%) and are accurate to quite close limits (typically within 3% of the actual voting).

The reason a 1% sample of Lit readers is not significant is that it is self-selecting. Only people with strong views will vote (apart from the relatively small number of habitual voters); this means that the extremes are over-represented, especially those with strong negative views.

In a crude voting system (like politics or Lit) it is difficult to work out why people are voting as they do. For example if Lit introduced compulsory voting then there would be an important group voting 1 on every story they read as a protest against compulsory voting, and another group voting 5 on every story for the same reason!

All good points. I knew that if I took this thread towards statistics that someone would jump in that knows more then I do. I know enough to be dangerous and that's about all. I did think that sample size and sample selection were important issues, I just never explained it as well as you did. Thanks.

 

[CorsetLvr]

That doesn't necessarily mean that though. You assume that there are just as many strong positive as strong negative views. However, if there are many more strong positive views, the strong negative views can be way overrepresented and still end up not influencing a vote as much as the positive views.

If there were as many strong positive as strong negative scores then everyone's score would end up as 3.0. (5X + 1X)/X = 3, where X equals an equivalent number of positive and negative votes.

 

[sr71plt]

That doesn't necessarily mean that though. You assume that there are just as many strong positive as strong negative views. However, if there are many more strong positive views, the strong negative views can be way overrepresented and still end up not influencing a vote as much as the positive views.

Sorry, I don't follow. First, "views" aren't "votes." You can't analyze "views," but you can do some analysis with "votes." As far as positive versus negative overrepresentation goes, when it takes three "5" votes to bring one "1" vote up to the 4 rating level, and the average vote across the board at Literotica is over 4, there's no way that this represents overrepresentation of negative votes. Logic and simple math.

It does mean, though, that a negative vote is more powerful (not that there are more negative votes than positive votes). Is that what you mean?

 

[fieryjen]

Okay, replace "views" with "votes".

I'm saying that people tend to vote positively more than negatively. My assumption is that while they are many people who vote 1, there are a lot more people who vote 5. You use the starting assumption that it's split evenly, which isn't necessarily true (in fact I'm fairly certain it's not).

For example, let's assume that there are two negative votes for every five positive votes. That would give a score of 3.8 for a story. Now, since we said that the negative votes are overrepresented, there should be more than just two for every five positive votes. Let's assume that there are twice as many. The score for the story would still be 3.2.

 

[sr71plt]

Okay, replace "views" with "votes".

I'm saying that people tend to vote positively more than negatively. My assumption is that while they are many people who vote 1, there are a lot more people who vote 5. You use the starting assumption that it's split evenly, which isn't necessarily true (in fact I'm fairly certain it's not).

For example, let's assume that there are two negative votes for every five positive votes. That would give a score of 3.8 for a story. Now, since we said that the negative votes are overrepresented, there should be more than just two for every five positive votes. Let's assume that there are twice as many. The score for the story would still be 3.2.

OK, thanks. We've said the same thing, then. Which means that negative votes aren't "overreprensented" here, since the across-the-board rating average is well over 4.

 

[Weird Harold]

I'm saying that people tend to vote positively more than negatively. My assumption is that while they are many people who vote 1, there are a lot more people who vote 5. You use the starting assumption that it's split evenly, which isn't necessarily true (in fact I'm fairly certain it's not).

A point to consider here: If a "legitimate" one vote is an indication of poor writing, poor story-telling, or poor spelling/grammar/formating, then Literotica's editoral approval policies should completely remove any "legitimate" one votes entirely, leaving a vote range from two to five for a median of 3.5 if votes are equally distributed over the full arange. (instead of the 3.0 median of the full 1-5 vote range.

Another point to consider -- one that sr71 touched on -- is voter motivation: the typical ratio of votes to views suggests that people only vote on stories that affected them strongly either positively or negatively and consequently something on the order of 60-75% of votes are either ones or fives with most of the remainder being twos and fours. Very few readers are "strongly affected" by mediocrity so a Three vote is almost always that rare and precious thing, an honest, uninflated evaluation of your story.

 

[snooper]

... Another point to consider ... is voter motivation ...

Does it surprise anyone that someone who doesn't like a story hits the [BACK] button, rather than go to the bottom and vote? Therefore many people only see the [VOTE] box if they like the story enough to finish it. Thus those who don't much like a story, but are not offended by it, probably don't vote.

And then there is the kind-hearted group of voters who only ever vote if they like a story. They don't mind giving a 4 to a story that lacks something, but they don't ever want to discourage someone by giving them a 3 or lower.

I suspect that these two factors cover the majority of the readers, because the average is so high.

The self-justification used by the person I know who never gives less than 4 is "Who am I to criticise? I doubt I could do any better." This is a fallacious position, and the error becomes obvious if you consider that few of us can play the violin, yet we know (by hearing them both) that a Menhuin recording is better playing than the kid next door who only started to play last month.