What Are 'Views'?

[KenNicottii]

There seems to be some confusion as to what a “view” implies, as measured by Lit. Based on my observations, a view appears to be a reader clicking on an individual page, not a story. Therefore when comparing two stories that have the same number of views, you have to consider how many pages each story has.

If one story is five pages long and one is ten pages long, and both have 10K views, then you can’t say that the stories have equivalent popularity. Yes, there are a lot of different ways a story is consumed, but that’s not the point I’m making. I know you have people that read it through front to back, others that quit in the middle, some that jump back and forth – or many other reading patterns. Once you get into the 10s of thousands of views, I think this all smooths out.

I’ve always looked at a metric I call ‘story views’: divide the views by page count. So in the example above, the 5-pager would have 2,000 story views while the 10-pager would have half as many, or 1,000 story views.

Thoughts?

 

[ElectricBlue]

There seems to be some confusion as to what a “view” implies, as measured by Lit. Based on my observations, a view appears to be a reader clicking on an individual page, not a story. Therefore when comparing two stories that have the same number of views, you have to consider how many pages each story has.

If one story is five pages long and one is ten pages long, and both have 10K views, then you can’t say that the stories have equivalent popularity. Yes, there are a lot of different ways a story is consumed, but that’s not the point I’m making. I know you have people that read it through front to back, others that quit in the middle, some that jump back and forth – or many other reading patterns. Once you get into the 10s of thousands of views, I think this all smooths out.

I’ve always looked at a metric I call ‘story views’: divide the views by page count. So in the example above, the 5-pager would have 2,000 story views while the 10-pager would have half as many, or 1,000 story views.

Thoughts?

A view is a click into a story, it's not based on pages - from the checks I've done.

I was curious myself on this, early on, so I experimented on my own stories - choosing a multi-page story, I could see the view count click up by one each time I opened it, regardless how many pages were in the story.

Dividing by page count is therefore irrelevant. All a view tells us is someone opened a story, it does not tell us they finished it. They could have back-clicked after the first paragraph, for all we know.

 

[SimonDoom]

I believe EB is right. The number of pages is irrelevant. Views are calculated by the number of times the story is clicked on. Each time a reader clicks on a story, it counts as a view.

 

[TxRad]

According to one self proclaimed number one author, a view is a sale. I'm not sure how he calculated the profits on a free site though.

 

[KenNicottii]

I believe EB is right. The number of pages is irrelevant. Views are calculated by the number of times the story is clicked on. Each time a reader clicks on a story, it counts as a view.

This doesn't match my experience. For example, two of my recent stories (The Flume at Falcon Rock and Trailing Along) illustrate this. Both were contest entries and are in the same category (EC), both have roughly the same score, and both finished almost identically in the contests. Yes, Trailing has been out for 21 months vs 14 for Flume, but I think that difference is moot at this point since these stories have probably gotten close to 100% of their expected readership by now.

Both stories have almost the same number of views (58K-60K), but Flume is 11 pp vs 6 pp for Trailing. It seems inconceivable that a story almost twice as long has gotten the same readership.

If you compare the other metrics, Trailing has 30% more votes, 50% more favorites, and 50% more comments. If you divide the views by the pages, then Trailing has nearly 80% more 'story views' than Flume - this seems more consistent with the other metrics.

I've also gone back and read an old 5-pp story that rarely has any additional views. The next day I saw the view count go up by 6. And this was when my last submission was months before, so I wouldn't expect viewing triggered by a new story of mine.

 

[ElectricBlue]

I've also gone back and read an old 5-pp story that rarely has any additional views. The next day I saw the view count go up by 6. And this was when my last submission was months before, so I wouldn't expect viewing triggered by a new story of mine.

Six views, 5 pages? There's the flaw in your argument, right there.

My test is evidence based, three pages read, five pages read, didn't matter. The view count goes up by one. What can I say? You're drawing conclusions from suppositions. Go do an actual experiment, and see for yourself.

What do you mean, you wouldn't expect a new story to trigger views of an old story? I publish a new story, my back catalogue sees an increase in activity, every time. People read a new one and like it, go read the back catalogue. Why is that a surprise?

 

[Bramblethorn]

This doesn't match my experience. For example, two of my recent stories (The Flume at Falcon Rock and Trailing Along) illustrate this. Both were contest entries and are in the same category (EC), both have roughly the same score, and both finished almost identically in the contests. Yes, Trailing has been out for 21 months vs 14 for Flume, but I think that difference is moot at this point since these stories have probably gotten close to 100% of their expected readership by now.

Both stories have almost the same number of views (58K-60K), but Flume is 11 pp vs 6 pp for Trailing. It seems inconceivable that a story almost twice as long has gotten the same readership.

If you compare the other metrics, Trailing has 30% more votes, 50% more favorites, and 50% more comments. If you divide the views by the pages, then Trailing has nearly 80% more 'story views' than Flume - this seems more consistent with the other metrics.

I've also gone back and read an old 5-pp story that rarely has any additional views. The next day I saw the view count go up by 6. And this was when my last submission was months before, so I wouldn't expect viewing triggered by a new story of mine.

AFAIK, readers don't see the page count until after they've already clicked on the story, which counts as a view, so there's no reason length should affect how many people click. I don't see anything in what you've described that's inconsistent with "views are clicking into the story, not individual page reads".

 

[Zeb_Carter]

A view is a click or tap on the the link in the sites story list. That's it.

Click on the link, it's one view.

Click back and click on the link again, it is another view.

Click back, then the link enough times and you to could have a million views.

Views mean nothing except someone clicked on the story link. It doesn't mean they even read the story. Only that they clicked the link.

Shocking, isn't it?

 

[nosebone]

What do you mean, you wouldn't expect a new story to trigger views of an old story? I publish a new story, my back catalogue sees an increase in activity, every time. People read a new one and like it, go read the back catalogue. Why is that a surprise?

I agree with this completely EB, I've had stories that languished with lower scores that became some of my higher scored stuff as a result of my subsequent work.


To the OP, views are an interesting metric, but I don't think you can really tell much from it. I usually just pay attention to the story's votes and overall rating. That tells me much more than how many people opened it (and maybe quit three paragraphs in).

 

[SimonDoom]

This doesn't match my experience. For example, two of my recent stories (The Flume at Falcon Rock and Trailing Along) illustrate this. Both were contest entries and are in the same category (EC), both have roughly the same score, and both finished almost identically in the contests. Yes, Trailing has been out for 21 months vs 14 for Flume, but I think that difference is moot at this point since these stories have probably gotten close to 100% of their expected readership by now.

Both stories have almost the same number of views (58K-60K), but Flume is 11 pp vs 6 pp for Trailing. It seems inconceivable that a story almost twice as long has gotten the same readership.

If you compare the other metrics, Trailing has 30% more votes, 50% more favorites, and 50% more comments. If you divide the views by the pages, then Trailing has nearly 80% more 'story views' than Flume - this seems more consistent with the other metrics.

I've also gone back and read an old 5-pp story that rarely has any additional views. The next day I saw the view count go up by 6. And this was when my last submission was months before, so I wouldn't expect viewing triggered by a new story of mine.

Your observations support the fact that page views don't increase the view tally.

It makes no difference how long a story is for view purposes, because when the reader initially clicks on the story to open it, the reader has no idea how long it is. It isn't necessary to read the whole story to count as a view; opening the story does the trick.

Short stories naturally have higher vote to view ratios, because readers are more likely to get to the end of short stories than long ones. I notice that there is a clear relationship between the length of my stories and the vote to view ratio. That would NOT be the case if viewing a page counted as a view. If you look at your stories, I'll bet they bear out this relationship.

 

[NotWise]

Short stories naturally have higher vote to view ratios, because readers are more likely to get to the end of short stories than long ones. I notice that there is a clear relationship between the length of my stories and the vote to view ratio. That would NOT be the case if viewing a page counted as a view. If you look at your stories, I'll bet they bear out this relationship.

That relationship between story length of votes/view is logical, but there's always something that breaks the simple trends. It doesn't hold true for my stories.

The longest story I've published has my second highest ratio of votes/views, but the length would indicate that it should have a low ratio. Another of my longest stories has the 4th highest votes/views. But then, another of my longer stories has my lowest ratio of votes/views.

My shortest story is in the middle of the pack as far as that ratio goes. My story with the highest ratio of votes/views actually is fairly short.

There are reasons other than the length that explain those ratios, and that tends to bust the simple relationship.

 

[RejectReality]

I'll say that my own tests years ago, performed on stories in low-read categories, which were a couple of years old, unconnected to any recently released stories, and in the wee hours of the morning, yielded results of one view, regardless of how many subsequent pages of the same story were opened.

These stories were getting minimal views per day at the time. I'm fairly certain it was single digits.

I repeated this experiment on multiple stories of similar criteria, with varying numbers of pages, ( Between 1 and 6 Lit pages ) on multiple subsequent days, and always came up with the same results.

At the time, views were real-time updated in the control panel. They only update about once every 15 minutes now. It was far easier to see the results then, and they were definitive.

 

[ElectricBlue]

At the time, views were real-time updated in the control panel. They only update about once every 15 minutes now. It was far easier to see the results then, and they were definitive.

We probably both did the same tests in response to another counter theory thread a few years back.

I'd forgotten the real-time update, but yes, that was clear to see.

The other test is the mathematical one - supposing the page count theory was correct, then Views divided by Page Count would be a whole number every time, which of course, it won't be, not every time.

 

[SimonDoom]

That relationship between story length of votes/view is logical, but there's always something that breaks the simple trends. It doesn't hold true for my stories.

The longest story I've published has my second highest ratio of votes/views, but the length would indicate that it should have a low ratio. Another of my longest stories has the 4th highest votes/views. But then, another of my longer stories has my lowest ratio of votes/views.

My shortest story is in the middle of the pack as far as that ratio goes. My story with the highest ratio of votes/views actually is fairly short.

There are reasons other than the length that explain those ratios, and that tends to bust the simple relationship.

The fact that story length is not the ONLY factor that determines the vote to view ratio doesn't mean it isn't significant. That was my only point. Ken observed that his shorter stories had relatively higher vote counts, and that fact is consistent with my hypotheses and does not support his hypothesis that views are based on page, rather than story, views.

You have written enough stories that I'll bet if you calculated the vote:view ratio for all of them and compared them to story length you would see a statistically significant relationship overall, even if not all stories conformed to the rule. Since you've written stories in many different categories it's not surprising you have lots of nonconforming results.

I analyzed all 21 of my stories, and found that there is a statistically significant, though not perfect, relationship between the length of the story and the view to vote ratio.

 

[KenNicottii]

I've also gone back and read an old 5-pp story that rarely has any additional views. The next day I saw the view count go up by 6. And this was when my last submission was months before, so I wouldn't expect viewing triggered by a new story of mine.

I worded that last sentence poorly – of course dropping a new story will (hopefully) bump the metrics of older stories as readers look to see if your other stuff is worth checking out. It happens each time I put a new story up.

What I was trying to get at was my page view experiment on the old story was done when it had been many months since I had posted a new story, hence there wasn’t likely to be any ‘background noise’ from other people clicking on any old stories at the same time I was trying to run the experiment.

Yes, I have no way of knowing if one or more other users decided to read the old story on the exact same day as I was trying to run the experiment. As far as the discrepancy between 5 pages and 6 views, I can’t be absolutely certain that I didn’t inadvertently add an uncounted page view somewhere in the process – or maybe the Lit software tallies for these metrics aren’t being handled precisely.

I will wait a bit longer to redo the experiment – I’m still getting spillover action from the story I posted 2 months ago.

 

[NotWise]

You have written enough stories that I'll bet if you calculated the vote:view ratio for all of them and compared them to story length you would see a statistically significant relationship overall, even if not all stories conformed to the rule. Since you've written stories in many different categories it's not surprising you have lots of nonconforming results.

I checked the correlation of votes/thousand views to story length in words. It came out to 0.27. I've read social and medical studies that went on to show that such low correlations were statistically significant, but I'll assert that it isn't. To the extent that there is a correlation, the relationship is for votes/thousand views to increase with story length -- the correlation is positive.

The category is certainly part of the reason for the variation. Another (and maybe the largest) is simply the way the story ends. My stories that end with an emotional impact get a higher number of votes/view. My highest votes/thousand views (22) goes to "The Third Ring," which is an epic tragedy.

Stories that are positioned to get disinterested views (wrong category, for instance) may have a very low ratio of votes/thousand views. My example is part 3 of "A Valentines Day Mess," which spent a few months on top of the I/T 12-month top list. It got a lot of views from people who where curious but had no other interest in the story. It's ratio is 3.3 votes/thousand views -- my lowest.

 

[SimonDoom]

I checked the correlation of votes/thousand views to story length in words. It came out to 0.27. I've read social and medical studies that went on to show that such low correlations were statistically significant, but I'll assert that it isn't. To the extent that there is a correlation, the relationship is for votes/thousand views to increase with story length -- the correlation is positive.

The category is certainly part of the reason for the variation. Another (and maybe the largest) is simply the way the story ends. My stories that end with an emotional impact get a higher number of votes/view. My highest votes/thousand views (22) goes to "The Third Ring," which is an epic tragedy.

Stories that are positioned to get disinterested views (wrong category, for instance) may have a very low ratio of votes/thousand views. My example is part 3 of "A Valentines Day Mess," which spent a few months on top of the I/T 12-month top list. It got a lot of views from people who where curious but had no other interest in the story. It's ratio is 3.3 votes/thousand views -- my lowest.

That's interesting that it's positive. I suspect it's because of the small sample size and the variety of different kinds of stories -- other variables are more important than story length.

For my 21 stories the correlation is .466 -- not overwhelming, and sometimes drowned out by other factors, but still somewhat significant. One of the factors that makes it hard to tell is that most of my stories are 3-4 pages; only 1 story has 1 page, and only 2 have as many as 6 pages.

It stands to reason that the longer the story is, the less likely the reader is to get to the end, and whether or not the reader gets to the end is a factor in whether or not the reader will vote. But there are a lot of other factors.

 

[NotWise]

I just did the test. I opened one of my stories that is eight pages long and clicked through each page. Then I waited for the Works page to update, which took a few minutes. When it did update it incremented the views by one, not eight.

For my 21 stories the correlation is .466 -- not overwhelming, and sometimes drowned out by other factors, but still somewhat significant. One of the factors that makes it hard to tell is that most of my stories are 3-4 pages; only 1 story has 1 page, and only 2 have as many as 6 pages.

I have a similar pattern. Many of my stories are just under or just over two full Lit pages. In that group the votes/view ratio is spread over the whole range that I see in my other stories.

 

[oggbashan]

A view could be a scan by a bot, not a human.

 

[SimonDoom]

A view could be a scan by a bot, not a human.

Undoubtedly true. In fact, it's clearly true regarding the number 1 story on the 12 month most-viewed toplist, , which has over 574,000 "views."

But I don't think this is true for most stories, certainly not to this degree.

 

[LightningSeed]

How bored would you have to be to want to set up a bot to view your story?

 

[SimonDoom]

How bored would you have to be to want to set up a bot to view your story?

Perhaps vain rather than bored, because with the right effort you can make your story look like the most popular story on the site.

 

[LightningSeed]

Fair enough. Fair enough.

 

[KenNicottii]

Hey, thanks to all the fellow authors who weighed in on this topic. My output is so meager (6 stories), that I couldn’t run large sample analyses like some of you can. But I did have that pair of stories that unusually had a lot in common (category, score, views, older than 1 year, finished top 10 in different contests). And they were both longer stories, at least twice as long as the supposed Lit sweet spot of 3 pages/10K words.

It seemed counterintuitive (at first glance) that the readership was essentially identical for 2 stories that were that dissimilar in length. And I was taking the more modest view that my readership might be inflated by page views.

I guess it’s unknowable how many of those views are from people who opened the story but immediately bailed when they saw the page count. Or were bots doing whatever they were up to.

 

[oggbashan]

Some of my stories have been stolen from Literotica by bots who just want content for their malware-infested sites. They never take more than the first Lit page, and don't worry about whether the story fits their site's title.

Why would anyone look for my mature romance stories on a site for 'Japanese anal-fisting teens'? But my stories are there.