View counts down?

[ColtonWrites]

Because the thread has faded out has new messages and I had nothing better to do than to muck rake, here's a screen shot of the most-recent 25% or so of the daily view traffic on my highest-view story.

I posted the story on 3/9, so this is long after it dropped of the New and 30 Days Top lists.

This is the rolled-up data for each day. I was trying to get it to display the hourly dataset (which would highlight all the zero-view data points), but I got sick of playing with it.

View attachment 2580471

Is that done in Excel or did you script it?

The data scientist in me wants to do something with my own data here and on AO3. Any suggestions on where to start with the data gathering? What kind of scraper are you using?

 

[designatedvictim]

Is that done in Excel or did you script it?

The data scientist in me wants to do something with my own data here and on AO3. Any suggestions on where to start with the data gathering? What kind of scraper are you using?

I should probably write a FAQ.

Been getting a bit of interest in it lately.

A Windows DOS batch script to do the collecting can be found in this thread.

That just gets you the raw data.

Earlier in that thread is a PHP script which does the data collection and dropping the data into a table. This earlier version is posted as more of a generic guideline than a detailed program, since I have no idea what or if anyone else uses a dev framework. But all of the pieces you'd need to roll your own should be there.

I have more parts of the PHP app (not directly connected to data collection) that does browser-based reporting using the Google charting library.

That's what I use to get the simple, clean-looking graphs.

Those graphs are just a browser-click away.

 

[lovecraft68]

This thread reminds me of something that used to be discussed more often in the past and that is the questions of what constitutes a view and how accurate are they?

You click your story its a view, someone clicks it, quickly decides it's not for them and drops off is a view but not a read. Bots, web crawlers, spammers leaving comments that we may never see because the site boots them, but I believe the view remains.

Someone clicking the story to just one bomb it is a view.

When you think of it in those terms, then the conversation about how poor the comment to view ratio also changes because some of those views are not 'real' and aren't readers.

As this thread shows, many people focus-almost obsessively in some cases-over this stat and it's the vaguest of all the stats out there. Comments, favs, votes, reader lists etc are all valid and provable.

10k views...but how many read the story? No way to know.

 

[designatedvictim]

This thread reminds me of something that used to be discussed more often in the past and that is the questions of what constitutes a view and how accurate are they?

You click your story its a view, someone clicks it, quickly decides it's not for them and drops off is a view but not a read. Bots, web crawlers, spammers leaving comments that we may never see because the site boots them, but I believe the view remains.

Someone clicking the story to just one bomb it is a view.

When you think of it in those terms, then the conversation about how poor the comment to view ratio also changes because some of those views are not 'real' and aren't readers.

As this thread shows, many people focus-almost obsessively in some cases-over this stat and it's the vaguest of all the stats out there. Comments, favs, votes, reader lists etc are all valid and provable.

10k views...but how many read the story? No way to know.

I'd seen some arguments previously on people suggesting that Favorites are a good metric.

On its surface, I kind of agreed at the time.

That lasted just as long as it took me to click out to someone who'd favorited one of my bits and saw that he had over five thousand favorites.

That kinda dulled the novelty of the concept.

 

[lovecraft68]

I'd seen some arguments previously on people suggesting that Favorites are a good metric.

On its surface, I kind of agreed at the time.

That lasted just as long as it took me to click out to someone who'd favorited one of my bits and saw that he had over five thousand favorites.

That kinda dulled the novelty of the concept.

Merit isn't really what I meant. I'm saying it can be quantified. So and so faved you or your story, its an acual favorite, whereas the view stats is simply a clicker and you have no idea if it was a real person let alone did they read the story.

You can pick apart a lot of stats here. Votes? People vote on their story, their friend in the forum run and vote on it for them, some people have several alts that they will use to vote on them. This may not matter much in a category like I/T or LW that has a lot of traffic but it does on low read categories.

Comments? Same, a major member of this forum and prolific author was caught-by me-commenting on their own stories with another ID, going on and on fawning over himself. Other people have alts that they use to do this. Then there are the authors who reply to every single comment. Statistically, the story has 40 comments when you see the totals, then you see 20 are from the author. Those aren't real comments as in from a reader, but they count in the overall stat.

There's been enough conversations about scores here, the high fiving cliques, the trolls, the insane category of LW where good stories mostly go there to die thanks to incels and cold male insecurity.

So, what's left that we can say "This is legit!"

Not a whole hell of a lot unless you take it all at face value and don't dwell on it.

Going back to the discussion in another thread, I think this is where coming to the forums can be a negative experience, you see all these discussions and get a crappy wake up call that the numbers your proud of may not be as real as you'd like them to be.

 

[RejectReality]

One last time. 2006 views and 11 votes. At least next time I go to look, I'll have a weeks worth of data to compare to.

 

[ElectricBlue]

I'd seen some arguments previously on people suggesting that Favorites are a good metric.

On its surface, I kind of agreed at the time.

That lasted just as long as it took me to click out to someone who'd favorited one of my bits and saw that he had over five thousand favorites.

That kinda dulled the novelty of the concept.

Across my story file (130 or so chapters or stories) Favorites are roughly 20% the number of Votes registered. My metric is one Vote per hundred Views, so Favorites is one per five-hundred Views. That's so far down into the noise as to be meaningless, especially since so many are clearly book-marks - like your 5000 guy.

 

[Bazzle]

Since my millionth reader, I have had ~4000.

So I am still getting plenty of clicks. No favourites or comments, but clicks.

 

[HeyAll]

The other day there was a new mom/son car ride story which got almost 60,000 views and 100 favorites in a 24 hour period. Old school numbers.

Big hits are still possible. But much, much harder to achieve given the amount of other stories. Sticking to familiar formulas is what works.

 

[MillieDynamite]

I haven't even looked at my view counts for a long time. Or at my control panel for over a week.

 

[Smuttyandfun]

The views seem to have stalled again.

 

[RejectReality]

Not here. I checked the two I always look at when I saw this, went about my day, and then came back to see 70+ views on the one and 20 on the other.

 

[SimonDoom]

My story views over the last ten days are back to average levels.

 

[SmilingLez]

As you can see there's a bug in the view count. I have 16 ratings and only six views. Quite impossible. Yes I put in a bug report, seems like if the title is numbers the view count is way off. The 25 plus year CMS is riddled with bugs

 

[Smuttyandfun]

The views have stalled again. My latest story just came out a few hours ago, it has 1002 reads. It's been stuck there for over a half an hour. Same with my most read story, and my story that came out a week ago. Once it starts again, I don't think they ever catch up with the views. Seems like the same things happens every night, though it also happened at 9:00am yesterday.

 

[THBGato]

Apologies if somebody already made this point, but how much of this is due to the change in the category hubs?

On the new "mobile friendly version" there are 10 new stories visible; on the "classic view" it's 25. Kind of recently, the site changed it so the new version is the default and you have to actively click back to the Classic View.

We all know that being on the "new" list drives views, so if the new list is 40% the size of the classic one that is going to hit views...

(I also don't like how the new version doesn't show recent comments. I often found stories that way. I also think that readers will have even less motivation to comment if they think that nobody will see their comments.)