Pixelfed user count has gone vertical.
Editing to let people know that I will be blocking anyone who feels the need to tell me why this graph is inaccurate. I truly don't care, but feel free to chime in with your useless take and land a spot on my block list! š
You're viewing a single thread.
Lies, damn lies, and graphs that don't have the Y-axis starting at 0.
10% growth in a day is nice, but far from a revolution. Let's see this trend going for a month.
Monthly active users increased by 43% between 13 and 14 January: https://pixelfed.fediverse.observer/dailystats
I don't think anyone here is arguing that the entire world will be using pixelfed by the end of the year, and that its usage will expand to other galaxies by the end of the decade.
It's a comment about the current growth curve, and it is both accurate and interesting.
Lemmy had the same jump in numbers during the Reddit Exodus. Mastodon had a huge boost when Elon bought Twitter.
Every spike has been a followed by a slide back to baseline in less than a couple of months. After you've seen it happen so many times, it is no longer interesting.
The fact that you believe these platforms were the same before and after these events makes it sound like you were not, in fact, there to see it happen. In my experience, it permanently changed both platforms, transforming them from weird niche sites to genuine alternatives.
That said, what you find interesting or not is not any of my business.
This thread (and others in that post) might interest you: https://communick.news/post/2320430/4138857
Yeah, I know I shouldn't bother. I am just annoyed by the misconception that all graphs should always start in 0 on the Y axis, as if it was some law of nature. Shouldn't allow myself to get dragged in further. :)
Good mindset :)
I am here since before the Reddit backout and I am on Mastodon since 2018. Lemmy was at 15k MAU, went up to over 125k and now is 1/3 of that. Mastodon had 1M 575k something before Elon, hit up close to 2M 1.5M and now is sitting around 800k. (edit: I was looking at the overall charts and used wrong figures. Corrected now.)
Sure, if your reference point is waaaay before the spikes then what we have now seem "a lot". However, my point is that these spikes are far from being indicative of mass adoption.
Lemmy was at 15k MAU, went up to over 125k and now is 1/3 of that.
So it increased by 200%
@porousgreymatter @rglullis reddit exit ?
Oh, wow, very impressive! Did you have to use a calculator to get to this challenging result?
Communick's revenue grew 1800% in 2024, compared to 2023. Do you think that makes it successful in any way?
Mastodon had 1M something before Elon, hit up close to 2M and now is sitting around 800k.
This stat is really surprising. Do you by any chance have a link where I can see this data?
I'm sorry, I think I was looking at overall numbers and not just Mastodon. According to https://mastodon.fediverse.observer/stats&months=48, it seems Mastodon was at ~575k MAU before Musk, shot up to ~1.5M early 2023 and is now sitting around 800k.
No worries, thanks for the link to the data, I appreciate it!
I don't think that either Mastodon or Lemmy slid back anywhere near as far as back to baseline? Sure, usage went down, perhaps even significantly compared to the peaks, but I think that both retained a lot more users than they had before their respective spikes. I'm an example of someone who came into Mastodon with the Twitter exodus and into Lemmy with the Reddit exodus, and I've stayed for both.
I don't mean that the numbers went exactly back where they were. I mean that every spike was followed by a steady decline.
Compare it with Bluesky now, or compare it with Reddit during Digg's meltdown. Their growth curves will look like an S-curve, not this series of discrete jumps followed by 40-60% loss.
@rglullis@communick.news tbh we shouldn't expect the adoption curve of any Fediverse software to be somewhat similar to the ones of centralised social networks, since Fediverse completely misses the commercial aspect that encourage key users to stay in the platform easing communities to stick to it as well. My guess is that without the action of commercial dynamics, the situation wouldn't be so different from the jumps-and-losses moments we're used to
ok, that's a fair point. But then this whole talk about "going vertical" and "exponential growth" is useless, and the only thing that we could (perhaps) try to take out of these mass migration events is to ask ourselves "would we able to reduce churn in the Fediverse without compromising on any principles?"
In other worlds, does this mean that the only reason that the Fediverse is small is because it is not as addictive as the other social networks? Does this mean that leaving Instagram and coming to PixelFed is the same as quitting unhealthy ultraprocessed foods and realizing that when you switch to a healthy diet you simply don't eat as much at all?
And if any of this is true, shouldnĀ“t we change the effort from "leave Instagram and come to PixelFed" to "Leave Instagram and quit all social media"?
@rglullis@communick.news I think is far more complex than that, but this:
In other worlds, does this mean that the only reason that the Fediverse is small is because it is not as addictive as the other social networks?
may be kinda true. Just think about how much time you spend on Instagram (if you use it) by actually talking/socialising vs how much time you just consume passively contents and ads. Chances are that the Fediverse is where most of your actual interactions take place. In this sense I really liked your comparison with unhealty food, however if quitting bad food habits means usually to improve your life, quitting social networks often means to cut relations to communities you may heavily rely on, especially if you are a memeber of a marginalised community. There is a whole history about black community trying Mastodon during the Twitter migration and mostly ditching it because the platform wasn't so welcoming after all but mostly because the black community sticking to Twitter was just too important. Is only an example that shows that communities tend to stick together and move away mostly for cultural and historical reasons.
So as Gollum loved and hated the ring, communities hate and love their platforms, but most of all they love density. I think the Fediverse have a chance only if we stop using the commercial platforms as a paradigm and even if many users ditch the Fediverse after the first try, the most motivated of them choose to stay because the genuinely like it, and they may be enohght to form a dense enought user base that can motivate communities to form and stick here
Got it, thanks.
Cool. You... don't have to engage in discussions that bore you. Why waste your time?
the head dev Dan Sup mentioned the number of active users jumping from 6k to 30k. we'll see how it holds, but there is strength in numbers, people only stay if other people see what they post there. i have good hopes and really want to ditch instagram
Normalized graph for you:
Not a vertical line, at all. For sure cool, but very exaggerated with a dishonest graph.
Wake me up in 2 months when 80% of new users churned.
I've never been on IG but I'm strongly considering a pFed account. Am I churn or am I miniscule net-new?
And yeah, it's a hope that the rumoured meta toxicity is somehow magically not on pFed. I wanna see my nephew's designs and art but not the influencerati junk I fear is on the captive platform.
The y axis starting zero does not change the fact that it's exponentially growing right now. Filter that link posted below for 120 days and it is still a nearly straight vertical line of growth.
It's not exponentially growing. At best, the growth rate has exponentially increased. These are VERY different things.
It was exponentially growing, the platform would come screeching to a halt and crash.
Wait what? The whole point of exponential growth is that all itās derivatives/integrals are also exponential.
Jesus fucking Christ, you people are miserably pedantic. Nothing about my original post is wrong, so please kindly shut the fuck up with your "well achtually" takes.
Ah, the classic response to being proven wrong on the Internet. Have a hissy fit like a 5-year-old and call everyone pedantic etc. You were objectively wrong but there's no problem with that, everyone is wrong about stuff all the time, the difference is how you handle that. You either act like a mentally grown person and just go "oh, damn, I didn't know that, thanks for correcting me" or act like you've done. No one is out to say that pixelfed isn't growing, but you're misrepresenting facts and then being a fucking child about being told so.
Actually, pedantry is correcting people for things that didn't actually need correcting because everyone understood the meaning from the context, so there isn't any reason for him to thank you because you didn't actually do anything of value.
Username checks out.
I'm not calling everyone pedantic. I'm calling you and the other guy who chimed in with pedantic takes pedantic. You can claim you're right all you want, but nothing about the title of the post "Pixelfed user count has gone vertical" is inaccurate. Whether it's the user count or the user growth rate that is exponential is irrelevant. Nothing about my post is misrepresenting the fact that on the graph, the number of users is going nearly straight up, exactly as the title describes. Feel free to keep trying to correct me, but I know I'm right because I have eyes and can identify basic shapes. I am very willing to admit when I am wrong, but this is not one of those cases.
I'm not sure if people are pedantic or sarcastic at this point
They seem to be serious. The more concerning thing is that people are agreeing with them. I guess common sense is not so common.
I got the same criticism about a similar graph a while ago. Numbers are on the left, people can clearly see how big of a change the graph shows, and I have no way to present the graph differently as it's straight from the website, but people still want to argue about it
Tbf graphics like that are commonly used to be deceitful, because people don't actually check the numbers on the left.
But in this case I think it probably doesn't matter lol
And then tell you're throwing a tantrum when you get annoyed by it.
Numeracy is certainly not so common
You're the one that made a false claim with a biased graph...it's okay to be wrong. I'm wrong all the time. No need to be so hostile about it. You could even edit your OP and add in the graph I threw together that normalizes the data. It's still a significant bump in users, which is cool.
The primary reason I am getting hostile is because there is a chorus of dumbasses who think they're being clever by pointing out that graphs look different when you view them on different time scales. No shit Sherlock. It doesn't negate the fact that this is probably the most significant increase in users the app has ever seen, which is really all my post is saying. On the graph, the line is vertical. Even zooming out 4 months, it's still vertical. Maybe zooming out further will make it less vertical, but that is beside the point. It's still growing fast. Even on your graph you can see the shift. Getting called a liar and having people try to minimize a significant trend is not something I will just let slide off my back. Keep arguing all you want, you're wasting your time.
Ah, so you don't actually understand why you're being called out. It isn't the timescale that "the dumbasses" are calling you out on. It's the Y-axis.
If you cut off the bottom, you can make any graph look vertical. Still, it's not insignificant and it's definitely a turn which is cool, but this is why people are upset at your post title.
The problem is less about the chart and the fact that you are taking a jump from a very short interval and trying to pass it off as something completely unprecedented.
Getting called a liar and having people try to minimize a significant trend
"Lies, damn lies and statistics" is not about calling you a liar, but how people can selectively use different data points to present information that supports their thesis or confirms their biases. I wasn't calling you a liar, I am just disagreeing with you about this being "significant".
If this growth rate holds for the next two weeks, then I'll gladly change my tune and start talking about a trend. But so emphatically making projections out of one or two data points is a fool's errand.
You seem sensitive and very angry about it.
Sorry for getting annoyed when people needlessly correct others to make themselves seem smart.
Eh... It's important to not spread misinformation and correct it when you see it.
It's also important to just be ok with being wrong and get on with your life. No emotion needed.
I am very willing to admit when I'm wrong. However, the original post is not misinformation. The line on the graph is nearly vertical. That is objectively true. Even going back 120 days, the last few days are clearly vertical, and a sharp jump in users from previous time periods.
People just like to correct people to make themselves feel smart, and I have very little patience for that sort of thing. If someone makes a legitimate correction or proves to me that I am wrong, I am happy to be corrected, but needless and inaccurate corrections are just irritating.
If you're happy to be corrected, then let me please help.
Exponential growth is dependent on the original size, and is usually a multiple of that size. 10 going to 20 then going to 40 then going to 80 is exponential growth.
10 going to 20 then to 30 then to 40 is linear growth. The increments are not related to it's original size.
Going from 290k to 320k is barely a 10% increase. The way the graph is clipped LOOKS like an exponential line, but that's because it's clipped. If we showed the full data, 0 to 320k, it would not at all look exponential.
And even if we drop the mathematical meaning and go with the colloquial meaning, when people say exponential they usually mean doubling. So 290k to 580k.
People are calling you out because you exaggerated too hard. You called 10% growth exponential. That's just wrong on the face of it.
Now I will wait and see if you lied to me about being happy to be proven wrong.
Again, as mentioned several times, the original post says the line went vertical, which it just objectively did. You could change the timeline and it might not look so vertical, for example making it show every day or every few hours, but that would be less honest imo. Even zooming out 120 days, which is the maximum time frame I could find for this data (https://pixelfed.fediverse.observer/dailystats&days=120), it is still a straight vertical line, and very clearly the fastest it has grown for a while. That's not misinformation or exaggeration, it just simply isn't.
As far as exponentiality, the growth rate is absolutely exponential. From the 10th to the 12th, it grew by 2500, then from the 12th to the 14th it grew by 5000 users. Then from the 14th to the 16th it grew by 25000. So if you want to get really technical, it's not exponential because it's more than doubling each day, it's starting approach factorial growth. So I guess I will cede that technicality, but the point remains, if I was in a statistics course and someone told me to identify the type of trend that this graph represents, I would personally go with exponential. You're free to make your own interpretation of what sort of trend the graph shows, but you would be wrong. It's not my fault that you and others are conflating exponential number growth with exponential growth rate. I never claimed the raw number of users was growing exponentially, just that the rate of growth was exponential.
Now go ahead and call me a liar for stating facts.
Ahhh, I see the problem.
What you said sounded like "membership is growing exponentially", which is false.
What you intended was "The rate of membership growth is growing exponentially". Which is true.
These are two very different things. If you truly meant that second one in your comment, then all I can say is you chose some very strange wording for it. And I'm not the only one that misunderstood.
The only thing worse than misrepresentation of statistics is completely misguided criticism of statistical representation.
While we're at it, the X axis doesn't start at zero either.
It's the perfect combo of karma farming and "well, achtually" to make your blood boil. Good thing karma isn't a thing on here from what I can tell.
There's no karma on Lemmy
Did higher karma increase the boots of your upvote on reddit?
No, but there were some karma requirements on some subs. And you could resell high karma account to advertisers as they would seem trustful
