Alright Medium, the Stats Are Starting to Need Explaining

Like many other people I am dissatisfied with Medium’s stats, as a UI or a UX it is incredibly subpar, the only way I could ever suppose that this was the way it should be would be if Medium had a higher paid tier that got decent stats and thus the lousy stats were what the rest of us got to whet our thirst for real stats.

But evidently they don’t, at least Medium doesn’t advertise any, and so I have never been tempted to get any of the more expensive Medium accounts because the stuff they do offer does not seem at all useful.

I mean it is offensive to me that they don’t offer better stats with any of their accounts, just from a business perspective, it’s idiotic and I’m paying for Medium, that means I’m paying for a service run by idiots! Not a good look for me either.

Also, this is a capitalist society, by being such absolute shit at capitalism Medium makes me question the value of society, which normally is something I would have to do all on my own but now it has been thrust upon me by the incompetence of Medium at capitalist business.

Preamble done, I shall now go through the things that suck about Medium stats

Why do I suddenly have a big uptick of people reading my stuff?

This question is one of the most important questions for people to have answered on any site where they publish things. It is measuring a thing called “engagement”, Medium may have heard of this thing, they put the word engagement on things, but the evidence would suggest that they think it is something that happens between young people when they are in LUV.

Often when I start getting a lot of views of my stuff I have a strong assumption why, that it is because I have posted a particular article in one place and it is getting a lot of views. But as the term analytics refers to the data used to analyze things it would be nice if I could in fact analyze my way to this conclusion instead of just making assumptions.

But sometimes I have no idea whatsoever.

To help people with figuring out what is going on with their “engagement” many sites try to help their user through the stats, does Medium do this? Let’s see — here for example is something going on from Mar. 18

Great Medium, thank you for telling me how many Reads and Writes I have but what would be really useful is to

1. Show distribution of articles getting this traffic.
2. show distribution of referrers.
3. To further break down the time usage. Because often you post somewhere at a particular time and you find that is the best time for you to be posting things to get a response from that site.

How to do this. I suggest clicking on the line and opening up another panel that shows the information for the particular date.

To be clear for anyone who does not do web development, design or UX — that is not a super brilliant suggestion that Medium could be forgiven for not having thought of themselves.

0 views stories that have more than 0.

This is actually what prompted me to finally write this article I have been putting off for so damn long, because What the hell.

Look at the stats below, the 3rd story “Krisit Noem, Linguist, signs new deal to Publish Dictionary” says it has 0 views and 0 reads

SCREENSHOT 1, IN 1 TAB

What’s wrong with that you ask? Surely it must happen sometime. Well the thing is just a few days ago I looked at it and it had more than 0 views and reads. In fact looking at the same stats page in another tab shows me the following

SCREENSHOT 2, ANOTHER TAB

And it’s not because suddenly it got some views in between tabs opening, because the views that the article has all came from May.

Now if you don’t know anything about Medium’s stats you might think that means that the views and reads show on per month, and it has 0 reads and views in this month, but no, the reads and views are lifetime.

So then you might say well obviously, this is just a bug. And I agree, except that if it were a bug I would expect it to be 0 views on everything but it is only 0 views on that one article. Yes but still it obviously must be a bug, it doesn’t make sense otherwise. Sure, but the problem is that Medium’s stats UX sucks so badly that it doesn’t make sense, and if the UX doesn’t make sense it calls into question if the obvious bug is actually an obvious bug and not just more bad UX.

No matter how stupid and broken something obviously is in Medium’s stats implementation you can’t be sure if it’s really broken, or if they want it to be that way and it is functioning as intended.

Zero Is Not Zero and Other Medium Math Tricks

In the following images you can see stats for a story published earlier today when typing this, now this is behavior I am used to, and almost behavior you might expect, given Medium’s descriptions of how they collect stats etc. (hourly, daily batches)

But the thing is this story has now been published for a lot of hours. And the stats for the day have been going up. But the story has 0 views and 0 reads here. But surely you say, that makes sense, the lifetime stats are updated daily.

Sure — the lifetime chats are updated daily with a notice of this fact with a small text at the top of the lifetime stats informing you of the fact with above that a big graph that is supposedly updated hourly, at any rate a graph that will show movement during the day while the stats of stories do not update at all.

And if you open the story to read its stats. You can see it has 0 views, 0 reads 0% lifetime engagement, and a monthly views at the same time of 10 views.

I mean, the UX of the stats page where different components tell a completely different story is pretty damn irritating. I might be willing to pay to get better stats and be able to find out what stories are suddenly performing well and why, and for what markets, but the UX story here tells me it wouldn’t be worthwhile making the payment, because this is the UX of a company simply incapable of telling you things in a cogent manner.

The UX is stupid. Stupid UX is not always a problem, for example the UX for a Hunt ICE Agents Game doesn’t necessarily need to be smart because the purpose of that game is not to make you smarter, but the UX of something that is supposed to make you smarter, like a stats page, needs to be smart because otherwise on the meta level people get the idea that the company making it is comprised of idiots.

Imagine also being a developer who worked on this and having it on your resume, I mean I would just have to hope that the interviewers don’t use Medium because the kindest question I could expect would be “what the hell happened?”

I’m sure there is some underlying architectural reason why things are this way, but the fact is architectural reasons are not enough to excuse having bad UX and worthless features, because Architectures can be fixed if one cares about fixing the other things.

Seems Nobody ever views between Midnight and 8 in the morning

This is the case generally, unless you end up having a lot of views one day, then the next day will have views at the beginning of the day — for example here we have one day that has 262 views and 5 reads one day, and that goes on to 5 views and 0 reads at 8 in the morning the next day.

Now yes, before you go off sniggering any about how many views I got and how few reads and point me to an article like the following

which has the helpful suggestion that

Having a lot of views, but only a few reads is a bad indication of your stats (more on this later).

This happens when you do not write quality content or you use clickbait as your title.

Thanks Moral Arbiter! That’s helpful!!

Now of course I would be more than happy to change my content and or titles to ensure more reads but because Medium does not give me a good way of knowing what exactly is getting all the views on this day and next to no reads, it is rather difficult to do so.

Also that article has an interesting definition of views and reads, which has changed my perspective

Views are just simply the indication of how many times people see your post on the homepage and their feed, but they never click and read it.

Reads, on the other hand, indicate the number of people who actually read your story and spend their time reading it.

I had been under the misapprehension that views also included some time reading a story but deciding to leave, which would make sense given the suggestion to improve content. But aside from that I can understand also that views have to also take into account homepage and the feed, so you get a lot of views for content that people decide not to click on.

But anyway it is really weird to me that I get all the time zero views from midnight to are 9 in the morning. Unless I am getting 100s of views in which case it switches over and gives views all the time.

It is really rather anomalous behavior for anyone familiar with any other form of view tracking across the internet. I’ve never seen a traffic log where things went completely dead for that long. Sure I could understand if one or two days you got no views in that time. But basically it is a permanent fixture of the Medium stats that it seems that I get 0 views during those hours.

Which is actually even more weird than you would expect. I think, like most English writing, that Luminasticity’s normal reader profile is actually somewhere in the U.S, but my midnight to 4 in the morning is actually prime internet usage in the U.S. Obviously some of the writers featured here are in the U.S but I’m not and I maintain the Luminasticity Publication.

So when something on a site is anomalous in comparison to the rest of the world, and quite steady in behavior, I assume it is code defined.

I believe that Medium has some sort of track flag, that gets turned off for low performing publications or authors during the hours of the day that they are generally least active — so their midnight to 9. Views and Reads are sitting somewhere but have not been run and the data updated, or perhaps queries against that data are turned off during those hours and it just falls back to showing zero — unless the hours before midnight have been high performing enough that it makes sense to keep it operating real time (where real time is the hourly batching).

This might seem paranoid to someone reading this, but I think it is incredibly more paranoid that I should think that somehow, strangely and almost miraculously, the hours of the night in which I am most likely inactive do not have any sorts of views — contrary to my extensive experience in any internet statistics I have ever had to deal with.

This could be a very sensible optimization, but also some real irritating crap, given that Medium evidently hasn’t told anyone this happens (from what I can find) and anyway you can’t pay for it not to happen by buying the non-existent premium data plan.

Speaking of Money

This is all so irritatingly bad I have even thought maybe Illuminati Ganga should make some tools to track the actual statistics, basically by crawling the various pages and providing a better UX for Medium’s writers to know what is going on. But not sure if it would get approval because after all, if it did turn out to be profitable one might expect Medium to finally wise up and make some sort of solution.

Conclusion

This article was written by me, Agent 86, normally I just maintain Illuminati Ganga’s Medium account and put our writer’s stories and poems up and similar kinds of editorial duties, but this whole Medium stats has been pissing me off for so long now that I decided to also write something.

As you might assume I am not the only Medium user who has noticed that something is incredibly off about all this

In fact there is a whole genre of writing that seems to only exist on Medium the tropes of which revolve around wondering why the stats suck so badly.

a propos the idea of fixing this atrocious implementation of stats I refer to the following, who has evidently put in a lot of work on the subject.

Related Articles

As a general rule this is not the kind of thing Illuminati Ganga publishes or cares about, so we do not have a bunch of things that relate to it.

So instead here I will put some of our articles that have had a reasonable uptake, with some descriptions

1. 735 Views, 442 Reads, 60% Read ratio

This article is really a part of a longer series of articles, and itself summarizes a number of articles linked within it.

The best of these linked articles is the following

Which notes that in approximately half the book Jim is actually a freed slave.

2. 649 Views 536 Reads 83% Read ratio

This article was actually published in the Haven, normally most of Illuminati Ganga’s articles are published in Luminasticity but we do have a program trying to build an audience for our writers by publishing appropriate articles elsewhere.

3. 762 Views 333 Reads 44% Read ratio

An article written by Agent 99 who goes into strategies of time travel conflict, especially those with a more powerful opponent.

4. 155 Views 72 Reads 46%

An article discussing how our changing views of Racism and Anti-Racism can be thought of as analogous to changes in fashion, but also how it can render previous generations great works of anti-racism this generations contemptible works of racism.

5. 169 Views, 90 Reads, 53% Read ratio

What are the obligations of an author, an internet of things provider, or a streaming service when they sell you something.

6. 1K Views, 643 Reads 62% Read ratio

Another in the series of Articles on Mark Twain and Racism, this one on what makes The Adventures of Tom Sawyer a racist book, and it isn’t for the reasons you would expect.

7. 1.5K Views, 1.1K Reads, 72% Read ratio

Agent 99 with another time travel article, this one talking about ways you might polish up your cred among impressionable people by talking up some cool concerts you’ve been too.

I believe the technical term that is wanted here is “being a poser”

8. 190 Views, 83 Reads, 44% Read ratio

Our bid on why Netflix sucks. A staple of the Internet Article diet.

9. 226 Views, 121 Reads, 54% Read ratio

An entry in our ongoing series of articles regarding Rock and Roll hall of fame, this one had some interesting comments on how to determine if someone should actually be in the hall of fame or not, a question many critics have struggled with.

10. 1.1K Views, 470 Reads, 39% Read ratio

The concept of Narrative and how it affects the arts.

11. 525 Views, 275, Reads, 52% Read ratio

It seems somehow appropriate, to me at any rate, that an article about McCartney should be at pretty much 50%.

12. 314 Views, 196 Reads 62% Read ratio

Another money making and time travel article from Agent 99. It’s pretty mean spirited, probably that’s why the read ration outdoes the read ratio of Paul McCartney Mr. Nice Guy, but not as many views because who wants to read about piss when you can read about Paul McCartney.

13. 234 Views, 109 Reads, 47% Read ratio

A sort of disturbing horror fantasy story, written by Agent 71, who often does some strange things with language for fantasy stories.

Finally, I will conclude that gee, Illuminati Ganga’s poetry never seems to crack the 100 views level, and some really short poems not even cracking any reads.

Obviously the explanation is that Illuminati Ganga’s poetry sucks, especially if I ask Mark Laurence linked earlier in this article, but I suppose, given all the poetry I see on this site with tens of thousands of hits, no rhythmic or melodic sense, no style, and colorless descriptions of how the writer feels some undefinable hurt, I think it just might be that there is an audience mismatch.

Illuminati Ganga’s Poetry page

If you have read all the way down here and enjoyed my bad-mouthing Medium you might consider supporting Illuminati Ganga, lord knows Medium never will!