Numberfags' Wet Dream

[Anonymous No.]

See the picture for the numbers. Just look at the big table on top for the quickest results.
If you want a quick summary, most categories of vtubers are either stagnated or dropping in views, when comparing march to april; which is to be expected since the quarantines over covid are ending, so people have less time to spend on watching streamers now. The most alarming drop rate is from the ‘others’ categories, which has small corporate vtubers and indies. Twitch vtubers are also doing better growth than youtube ones, in a rough general, although they are still much smaller in their view/time rates.
I’m using the metric (vod) views/time because it seems to be the best metric we can use for comparisons. There’s no way to calculate direct hours watched without having access to data analytics from inside each channel owner or bullshit it; simply counting total views doesn’t take into account how much did they stream to get there (and they usually count small skits and music videos as well); and live viewers (on youtube) are very buggy and cut off a significant portion of that real number (for example, comparison between Miko’s twitch and youtube minecraft streams seem to indicate that 35% of her yt live viewers numbers are hidden [read the OP or the whole thread >>2253454 for anons discussion on the possible reasons for it).
Reading and trying to compare all those ratios for each hour length would have been a pain for any of you, so that’s why I’ve already gone ahead and made the ranking tables, with ranks for inside each time length. The longer the stream time, the fewer vtubers that had any videos available to calculate. Never compare general view/time ratios of streamers who have very different stream duration averages - that’s why I painted those general rankings on right in grey. Only compare data (or ranks) from the same time length. .
Through checking their ranks in the various time lengths, you can say that Haachama is in the sole top tier; closely followed by another with Pekora, Marine, Gura; then another with various such as Aqua, Ayame, Mito, Toya; then comes another tier with vtubers like Korone, Kuzuha, Shion, Rushia; then another with Helesta, Uiha, Suisei, Miko, Subaru, Noel, Shigure, Tamaki; and son on.
Also, I had a cap of 30 videos per streamer, so it might miss a few videos for a few chuubas, but it’s good enough to get their ratios already. Video types like short skits, short amvs/covers/original songs, members only, or private were not included for the calculations. Videos analyzed for 'this month' are from the time period of 19-03-21 to 18-04-21 (~10pm jst). Streams from this recent weekend were not included.
You can see the biggest growth fluctuations with those vtubers who had some huge special stream (like 3ds, returns). They can be big rises if the special vids happened on this month period (few examples: Flare, Mori, Suisei, Fumino, Lulu, Nyanners, Veibae); or big falls if the special vids occurred on last month (examples: Gura, Sora, Okayu, Roberu, Mayuzumi, MaiR). On those cases you can see % change of one specific time length way superior than the rest for that vtuber.
Collecting all this data was a whole lot of work, I’ll likely not be doing this again, unless some very significant event happens to threaten to turn this whole current trend around. I don’t think I missed anyone too big, I left out the many smaller nijisanjis, or holostars or whoever else because I don’t have the time to include them all - but I did include a few smaller vtubers I enjoy, just out of curiosity. If you think I left out someone relevant you can try convincing me to do their numbers as well, although if they are too small I’ll just tell you to fuck off.

[Anonymous No.]

Too fucking long, only read the first table. Nijikeks have a lot to explain now lol, didn't they love to boast that hololive was dropping hard? Were they so obsessed with hololive that they never looked at their own numbers?

[Anonymous No.]

>>2714355
Because they desperately want the crown of the king obviously. Same reason why Nyanners fans are obsessed with Gura in the EN space and created some kind of weird artificial rivalry which Gurafags took the bait on.

[Anonymous No.]

There's already a numberfag thread, retard. Learn to use the catalog next time.

[Anonymous No.]

>>2713927
Jesus fucking Christ mate, uses a bloody chart

[Anonymous No.]

>>2713927
Nijibros...

[Anonymous No.]

>>2714355

They only counted their top dogs and treat the low tier never exists , just to make the number look good on their side.

[Anonymous No.]

>>2713927
the two non holos that are listed as examples of biggest falls both had growth in their vod views over the last 30 days compared to previous 30 days?
https://socialblade.com/youtube/c/hoshinomea
https://socialblade.com/youtube/c/mayuzumix

[Anonymous No.]

>>2715243
>>2715243
>Video types like short skits, short amvs/covers/original songs, (...) were not included.
I only included real streams, short amvs can get a lot of views, but I did not count them.
Socialblade also seems to accumulate the views on any video available on the channel, while I only counted in the 1 month period. If for example theres a 3 month old video still getting a lot of views, it counted for them, but not for me.
If i didnt do this, all the idols who put out a lot of songs would be dominating it, artificially.

[Anonymous No.]

>>2713927
I ain't reading all that. But I'm happy for you. Or sorry that happened.

[Anonymous No.]

>>2715539
Seems a poor measure if it can call people who are objectively speaking growing some of the biggest decliners.

[Anonymous No.]

>>2713927
Please share the spreadsheet.

[Anonymous No.]

how tf have you done all this and not made a table of best/worst growth and or viewers per unit time what the fuck i just wanna see whos shit in holo and whos not

[Anonymous No.]

>>2715704
This.

Especially when in the write up you say that:
>Haachama is the sole top tier, closely followed by Pekora, Gura and Marine

Make a chart of this shit that summarizes everything

[Anonymous No.]

>>2714355
Niji has way more vtubers and most of them treat it as a hobby at best, of course their average is shit.

[Anonymous No.]

>>2713927
very interesting stats. Great job OP. Please share the spreadsheets

[Anonymous No.]

>>2715221
imagine if they added all the low tiers, their average would be much lower

[Anonymous No.]

>>2713927
thanks anon

[Anonymous No.]

>>2716108
Well there's the middle table, ranking them in each time range. I did not know how to make a total ranking counting all the time ranges, without it fucking up in either of these ways:
- just do avg of the rankings: no, you being 20/123 in the 1hr isnt the same as being 20/21 in the 9+hr. People like Astel would get a false boost in ranking.
- just rank the total view/time: no, as i said on OP, longer streams get smaller ratios for any tuber, because they get away from prime and into dead hours. I left that rank there, in gray, but do not use it, it means little (for example check Tamakis case: ranks 19 and 51, but because all the streams are short, this way would get them to 10th placevsomehow)
- invert for example 1st/123 to 1/123=0,00813, then avg and rank from lowest. That would seem fine, but for example even if someone had 1st in all ranges, them doing long streams so getting 1st/21=0,04762 in 9hr would fuck them up good and make them lose to, say, a vtuber who was only 3rd streaming only in 1hr: 3rd/123=0,2439, against that 1st in 1&9hr tiers with 0,02787, to give a rough example.
Give me a better suggestion to beat that and I'll do it.

spreadsheets link (csv):
https://www.sendspace.com/filegroup/kY2Oc%2BWuTK%2Fhp%2BuxEhdMVjRBuC0lxDcD
Theres some trash around and I never made it thinking another person would try to read it. Dont complain.

[Anonymous No.]

>>2713927
Daily reminder that numbers aren't real

[Anonymous No.]

..so this means haachama is the strongest idol?

[Anonymous No.]

>>2716108
>>2717302
since nobody gave a better idea, I did the last option, which seems by far the least bad

[Anonymous No.]

>>2718391
always has been

[Anonymous No.]

>>2720647
I'm honestly surprised Pikamee is that low, and right below artemis of all people

[Anonymous No.]

>>2720647
>Ina about to pass her in subs
>Moona about to pass her in views
>Mori catching up in scs
KFP can`t get one win, can they

[Anonymous No.]

>>2716390
Cope.

[Anonymous No.]

>>2713927
A hundred million viewed music video doesnt equate to a hundred million unique viewers,vod is fucking bs and doesnt need to be taken seriously no matter how much graph you make from mental gymnastics

[Anonymous No.]

>>2721764
Try reading the thread before posting. Short music videos were not included, only real streams.

[Anonymous No.]

>>2721764
>Video types like short skits, short amvs/covers/original songs, members only, or private were not included for the calculations
read nijikek

[Anonymous No.]

>>2721838
>>2721846
I set music vod as an example, im just saying that a person can click a video many times fucking numberfags

[Anonymous No.]

>>2721838
>>2721846
I think what he’s trying to say is that vod views isn’t a good metric since they don’t actually equate to unique views, which doesn’t really matter since the ones who have more people watch them will have more either way

[Anonymous No.]

>>2721927
>>2721963
>a person can click a video many times
>don’t actually equate to unique views
you have to watch at least 30 sec to get a YT view and even so there is a limit, around 5 I think, of how much views you can give to a video

[Anonymous No.]

>>2722333
It isn't a hard limit like that. It will stop counting repeated plays if you watch it a lot during short time periods to prevent abuse but then after a while it will count them again.

[Anonymous No.]

>>2722427
I guess thats why OP didnt include song or short videos

[Anonymous No.]

>>2722541
But if we are saying multiple plays are bad to such an extent that song videos should be discarded why is it then fine to ignore that the same thing will be effecting the figures for all videos? It's just as flawed a method of measuring anything as the live viewers measure it claims to be more reliable than.

[Anonymous No.]

>>2722809
>why is it then fine to ignore that the same thing will be effecting the figures for all videos?
unlike songs or short videos people dont click a stream more than once or twice

[Anonymous No.]

>>2722930
They definitely do. People leave streams and come back all the time and come back to visit vods to look for certain bits repeatedly.

[Anonymous No.]

>>2722930
Sometimes I go back to rewatch a vod or go to check something out from one a few times, but it is true that people click on music videos more

[Anonymous No.]

>>2721220
Good thing she only cares about her friends and not the fucking numbers!

[Anonymous No.]

>>2722809
>>2722952
About multiple views, sure it must happen, but I have been given absolutely no reason to believe it would happen so many more times with one vtuber than another, or that the repetition is even anything relevant to the total. It also counts all the late comers who miss the live and watch the vod instead

[Anonymous No.]

>>2722809
>live viewers measure it claims to be more reliable
how so? when there is a lot of people like me who wants to watch holos jp streams but I cant cause times zones

[Anonymous No.]

>>2713927
>invented a non sense method to fit your narrative
lol, do holofags really?

[Anonymous No.]

>>2723357
cope

[Anonymous No.]

>>2723187
I'm just saying why should we throw out the main long time accepted method of comparison on the grounds that it is inaccurate just to replace it with another method which is quite obviously flawed and inaccurate too?

[Anonymous No.]

>>2723513
>quite obviously flawed and inaccurate
you have no real arguments against it, only assumptions out of your head.
Now look at the picture on the left.

[Anonymous No.]

>>2723513
even if the VOD views as a metric is flawed, I think it still shows a more concrete image of what’s going on than live viewers as a metric does since the amount of the latter depends on a bunch of factors, but at the end of the day a vtuber will usually have the same number of people check out their VOD than a live stream

[Anonymous No.]

How come a video have x10 the live views numbers when it gets uploaded in youtube after the stream ends? Does that mean that 90k viewers use adblock? Please explainit to me without pointing to an extensive graph

[Anonymous No.]

>>2723790
Lots of people coming in and then leaving so the live view number doesn’t get effected much

[Anonymous No.]

>>2723636
I don't really think you can really even argue with it being inaccurate and flawed without seeming kind of retarded. The very reason song videos are being excluded even by the guy who did the stats is to try to control for some of that inaccuracy. But it still just ignores all the rest of it and doesn't control for it at all.

>>2723674
VOD views also depend on a bunch of factors and vary quite a bit from stream to stream. Pekora playing Minecraft for 3hrs 700k views, a few days later Pekora playing Bravely Default for 3hrs 300k views.

[Anonymous No.]

>>2723838
The only one retarded here is the person making claims without any real data to base it on. Every metric is flawed in some way, this one just happens to be less so than the others.
And to your second quote, what point are you even trying to make? More popular games get more views, on both live or vod, not shit?

[Anonymous No.]

>>2723838
>Pekora playing Minecraft for 3hrs 700k views, a few days later Pekora playing Bravely Default for 3hrs 300k views.
True, this is why I think it would be better to compare the amount of VOD views the same types of games bring, after a certain amount of time has passed to let views from people who missed the stream and can only watch the VOD later accumulate
At the end of the day this is all kind of pointless when vtuber numbers depend on so many factors

[Anonymous No.]

>>2724082
Oh, and you also have to take stream length into consideration

[Anonymous No.]

>>2713927
>no statistical analysis
>no CV%'s
Do it right or don't do it at all.

[Anonymous No.]

>>2723024
surely this must be a jest..

[Anonymous No.]

>>2724021
>The only one retarded here is the person making claims without any real data to base it on.
But that is exactly what you are doing? You are claiming this method is less flawed but have nothing to show that. Your graphs demonstrate anomalies in live viewer figures but do nothing to establish how that would compare to the inaccuracies introduced by multiple viewings.

>>2724082
>compare the amount of VOD views the same types of games bring
When you are getting that granular you might as well just say you can't read anything into any of it considering for a lot of people it would be a long time apart.

[Anonymous No.]

How long until Vtuber Numberology becomes a full-fledged science? OP pretty much wrote entire abstract, intro, methods, results and conclusion sections, even if summarized. Were he to textually expand on it, it could unironically be a thesis.

[Anonymous No.]

>>2724250
>nothing to show
are you blind?:
>>2723636
another picture on the left
and ill give you a free third one in just 2 minutes cap

meanwhile, you got nothing but weak rrats

[Anonymous No.]

>>2724393
There is absolutely nothing in either of these images showing that multiple viewings have less of an impact on the accuracy of the analysis than live viewer anomalies in any of these images anon.

[Anonymous No.]

>>2724393
>>2724250
picture on left are live views. Everybody suddenly drops hard on december, and in the following months its slow growths or declines depending on the chuuba.

[Anonymous No.]

>>2724447
You are still just showing inaccuracy in live viewers and doing nothing to establish that it is any worse than the inaccuracy in vod views despite claiming vod views are more accurate.

[Anonymous No.]

>>2724437
anonchama, when you make up something out of your ass, it is on you to prove me, not on me. If you say you are cow typing your posts, I dont need to do anything, its you who have to show prove to being one.

[Anonymous No.]

>>2724447
Much better, thank you! AZKi inclining hard!

[Anonymous No.]

>>2724572
I didn't make anything up though? I was merely questioning why we would replace one flawed measurement system with another that is obviously flawed as well. You on the other hand came out with the claim that it is less flawed with no data to back that up.

[Anonymous No.]

>>2724250
Is there a metric that you find decently reliable?

[Anonymous No.]

>>2724613
You havent shown a single proof of multiple views being a problem, because guess what? You dont have that data. i cant get that data either. Having no info on it doesnt mean it is anything relevant, nor that it isnt, it just means it can and should be ignored. Or I could just come up and say hey, theres a chance a bug is screwing up viewers from southwest usa not being counted. I cant prove it, you cant disprove it, it doesnt mean shit either way.

[Anonymous No.]

>>2724735
But the question remains why would we replace one flawed system with another flawed system when you can't demonstrate it is any kind of improvement? It doesn't logically follow that since one system has flaws the other one must be more accurate.

[Anonymous No.]

>>2724613
are you retarded?
>multiple images show that live viewers are flawed since December
>no images or real info show that vod views are flawed
so what can we conclude from this?

[Anonymous No.]

>>2724811
Not him, but the proof that using VOD views as a metric is flawed is in YouTube’s own criteria itself, you don’t need pictures for that

[Anonymous No.]

>>2724447
Not everyone has to have same growth, this just shows that gura cant keep her fanbase

[Anonymous No.]

>>2724811
Answer this question oh enlightened anon
>>2723790

[Anonymous No.]

>>2725107
>gura cant keep her fanbase
oh nobody is trying to deny that

[Anonymous No.]

>>2720936
Shartemis of the poo is actually very popular with (you)

[Anonymous No.]

>>2725166
>>2723822

[Anonymous No.]

Im just glad that my oshi is pekora

[Anonymous No.]

feels good man

[Anonymous No.]

>>2713927
>>2720647
top 20, not bad for a tako

[Anonymous No.]

>>2720647
GET REKT NOUSAGIS

CHAMMERS IS STRONGEST IDORU VCHUUBA

[Anonymous No.]

>>2724447
Ina has more viewers than Mori, Kiara or Watson despite having the least subs in HoloEN

[Anonymous No.]

Ina is the only HoloEn I enjoy watching.

[Anonymous No.]

Hello Numberfriends, I'm interested in the following statistic, so if any of you excel magicians could run a script on it would be much appreciated.

Ratio of superchat $ per hours of time streamed. (Could be either as a running total or something smaller like month-over-month, maybe both to see if there are any divergences and trends developing).
Basically an answer to the question of "how much $ gets the chuuba out of putting in one hour of streaming, on average"?, or basically their "hourly wage" in a way. So as to find out who really pulls those highclass escortgirl rates versus who'd be actually better off just flipping burgers for minwage etc.
(I know superchats are not their only source of income, but it's the only one publicly available and hey this is just for funsies anyway).


I'm mainly interested in just Hologirls for a start, but maybe there's also some insight in comparing other groups also.

[Anonymous No.]

>>2733980
Different anon here, with a different question.
Do superchat readings streams affect the superchat $ per hour, and if so, by how much?

[Anonymous No.]

Why won't the shark go back to minecraft to boost her numbers. Why is she playing games that filter her audience like Metroid and RDR2. Why doesn't she do more archived karaoke to boost her vod views and earn more SC. Why doesn't she make short meme videos. Why won't she

[Anonymous No.]

>>2740763
>your brain on numbers

[Anonymous No.]

>>2713927
Good stuff numbersbro, retards can fling shit at each other to their hearts content but at the end of the day the numbers don't lie.

You have any data or thoughts on clippers? Is it possible that casual watchers are foregoing streams in favor of short clips?

[Anonymous No.]

>>2734395
it does by a huge amount
>>2733980
I don't know how much value there is in this analysis. Because the distribution of superchats over time is not even and to a large extent independent of the amount of time spent streaming non-superchat reading content. A chuba can stream 5 hours of game content and the same chuba can stream 2 hour of game content. Both of those streams will have most of the superchats at the start and at the end of the stream and the overall difference won't be significant. Longer streams will naturally drag down the average.

[Anonymous No.]

>>2724371
You can numberfag for any business and big companies have people who it's their entire job to do numberfag reports.

[Anonymous No.]

>>2740763
Self Debuff so senpai can catch up...

[Anonymous No.]

>>2733980
Alright I did a little quick napkin calculation myself to get a first idea what these numbers might show. Total superchat amounts are taken from playbook, and for total video length (as a proxy for amount of hours streamed, without going through everything manually, so yeah the data is not as "clean" as I'd like but whatever) I went on each of the girls' youtube channel page, video tab -> uploads -> play all, which autocreates a playlist. Then I pasted the playlist link into this https://ytplaylist-len.herokuapp.com/ and noted the total time (converted into just hours).

First some semi random examples with HoloJP girls to demonstrate some sort of "types" that emerge (for example "low streamer /low earner", "high streamer /high earner", and every other combination).

Watame:
$ 598850 / 1411 hours
= $ 424,42 per hour

Ayame:
$ 442796 / 331 hours
= $ 1337,75 per hour

Mel:
$ 84839 / 246 hours
= $ 344,87 per hour

Rushia:
$ 1649030 / 818 hours
= $ 2015,93 per hour

(of course, as an additional next step you would ideally also factor in the inception date of the channel and adjust for that (maybe with something like "number of days of channel existence", and you divide the total video hours by that number, so you get something like "average hours streamed per day" and use that for the $ calculation), so you can compare the streaming hours more "like for like" instead of letting older channels naturally have more weight, but whatever this is just a first approximation for fun).

Anyways, some comments. Watame is a type "high streamer / (all relatively, of course) low earner". Ayame is "low streamer / high earner". Rushia is, I don't know if I'd put her in the high or low streamer bracket, perhaps medium maybe (again, here it would be handy to have a total overview of all Holos to compare, and adjust it for inception), but she is making bank in any case. And Mel is well, low on both counts, but again, this can't account for all the various reasons why that might be, involuntary absence for example, of course.

It would be interesting to do this for all holos, and graph them out on a scatterplot with a streaming and dollars X and Y axis to see the "type" groupings also visually.

As a bonus here's the five HoloEN girls. Because they have basically the same inception date, this also has the convenient feature of really being a more apples to apples comparison between each other.

Gura:
$ 408752 / 400 hours
= $ 1021,88 per hour

Ame:
$ 274899 / 634 hours
= $ 433,59 per hour

Ina:
$ 180325 / 420 hours
= $ 429,35 per hour

Mori:
$ 583397 / 399 hours
= $ 1462,15 per hour

Kiara:
$ 459573 / 684 hours
= $ 671,89 per hour

Maybe it's too early to group them into types and you need more time and data of course, but let's just comment on what we have. Mori is the extreme on both ends of the scale here, with both most dollars and least streamtime (although t the time is similar enough to Ina and Gura, so that's not really an insult), and clearly the most Money/Time-"Efficient" of the group. Again, a deeper analysis can of course argue that this also has much to do with her music, for which she also puts in additional hours offstream, so that might not be really "fair" enough.
Gura surprised me a little bit, because "chumbabbies are mostly kids, and don't donate as much" gets often thrown around, but for her relatively low streamtime her average donation numbers make her the second highest per-hour-earner. Ame and Kiara are both rather "high streamers" and put in the hours, but what separates them is the clear difference in SCs, almost double. Despite Ame arguably being the thread favorite, teamates don't really seem to actually put their money where their mouth is. Just kidding, but yeah, she has a relatively low ratio, for example compared to Ina who has streamed a third less time but netted basically the same hourly rate.

[Anonymous No.]

>>2740763
Reading this reminds me of this random nip who is obsessed about numberfagging Gura. https://twitter.com/nqve1

[Anonymous No.]

4 vtubers!

[Anonymous No.]

>>2742848
I assume this is Superchat earned at all time, because Mori (and by extension Amelia) should in the lead due to having a major event occur (Birthday Celebration).

I'm assuming Gura and Kiara to bump their numbers respectively when their Birthdays come up.

[Anonymous No.]

>>2742848
I assume this is Superchat earned of all time, which should explain why Mori (and by extension Amelia) being in the lead due to having a major event occur during the period (Birthday Celebration).

I'm assuming Gura and Kiara to bump their numbers respectively when their Birthdays come up.

[Anonymous No.]

>>2743729
Yeah, it's total SCs, since it's a conveniently displayed number on playboard, but of course I admit this would need further adjustment and refinements to approach something of a serious analysis. I'm basically just putting the idea out there so that any interested professional numberfriends can pick it up and develop it further if they're interested.

Oh and I just now realized I made a stupid babby mistake, the superchat amounts I took from playboard were initially displayed in €, not $ as I wrote down (so disregard that sign and just multiply by the current exchange rate)

Good point about the birthdays, Mori and Ame are the only ones that had their birthday streams yet and those are usually major moneymakers.
BTW I checked back out of curiosity and Mori made $55553 from her birthday stream, while Ame made $8934 on hers. Oh well...

For Mori, her birthday was her highest grossing superchat amount in a single stream, but for Ame her birthday didn't bring in as much and she had other streams with higher $$s (first monetization stream back in September got $21330, her stream as she reached 1mil subs got $15976, and her New Years stream pulled $12529). Mori's three highest streams, after her mentioned birthday stream, were her 1 mil subs stream ($44817), first monetization stream ($26983) and a "first free concert" stream in October ($16132).

For completeness' sake here are the numbers for the other members on their comparable event celebration (so far, monetization stream).

Gura made $21400 on her monetization stream, but that wasn't her highest grossing stream (which was beaten out by her 2mil subs stream at $23411).
Ina made $9838 on her monetization stream, but that number doesn't tell the whole truth, as she already had monetization enabled first on a normal game stream (call of cthulhu) and takos blew most of their load early ($16910 in that one, so "technically" $26748 in total for monetization?). Other than that, she had no single stream break above 10k anymore, but yeah she's not really the type to milk this stuff anyway. And Kiara's monetization stream netted her $9958, although she did a superchat reading right the next day for an additional $10783, so if you want to combine those two for what it's worth she also made around 20k off that event.

[Anonymous No.]

>>2744582
Thank you numberbro

Yeah, Amelia's 1st birthday celebration was sad. And I don't mean financially

[Anonymous No.]

>>2723187
It depends, i watch pekora's stream to the very end the first time i click it while during gura's , i just check on her stream many times just to get updated because she is unironically boring when nothing happens

[Anonymous No.]

>>2745152
Have you heard of not watching the stream if you're not entertained?

[Anonymous No.]

>>2713927
i like number but there's too much number
i dont like it

[Anonymous No.]

>views/time
How the fuck is this a good metric? What the fuck does the length of the stream have to do with views?

[Anonymous No.]

according to table, i trust vshojo is Vtuber new trend

[Anonymous No.]

>>2746030
>live viewers is not a good metric because unconfirmed Youtube "bug"
>VOD views is not a good metric because people can watch multiple times and be counted as multiple viewers

Views/Time is a good middle ground

[Anonymous No.]

>>2746626
I think the problem with simply total VOD views is the time they needed to reach that.
For example, which does better: a 1 hour stream with 100k views, or a 10 hour stream with 110k views? In this case, the first streamer is clearly performing better, but just simple views would tell you the second one is better.

[Anonymous No.]

>>2746835
but it doesn't tell you anything about the time after when it still accumulates views?

[Anonymous No.]

>>2746914
Im not sure I follow. Both cases would accumulate more views after time.

[Anonymous No.]

>>2746973
So the logic of accounting for time is that if they stream 3 hours they would get more views in that time than if they stream 1 hour holding all else equal. Time influences the number of views you would expect. The same remains true as time continues after the stream. It's not just the length of time during the stream that is influencing the view count but after aswell. Why account for the time during the stream but not after it. If dividing by time the stream ran then why not also count the time after and divide by that too?

[Anonymous No.]

>>2747125
have you ever made a youtube video?

[Anonymous No.]

>>2746973
I think the issue is that an older stream had also more time to accumulate views, so you have to take their dates into account when comparing vod views across different streams.
Or to illustrate it with the same above example - which did better, a video with 100k views that is a month old, or a video with 110k views that is 10 months old. Simply going by just the raw view numbers would tell you that the second one is better, but it's arguably the first one, because it reached its number in just one month whereas the second video was available for ten times longer (and despite that has just such a meager lead).

To account for that, you should adjust views by datelength. For example, divide the 100k views by 1 month, and the 110k views by 10 months (to get something like "average view per month of availability", which would be 100k for video A and 11k for video B, making it much clearer that A is the "better" one (in terms of performance numbers)

[Anonymous No.]

>>2747259
OP here, I think you're talking about a different matter but just to clarify: on my list of videos there was only videos from at least 2 days (collected late sunday, only from friday or prior) to a month old (the cut).
Another thing, i had collected the times of the videos in two parts, first half and second half, and collected the views for the first half, two weeks prior... which I had to update it all again, because im a dumbass and wasted that time.
Anyways, what I saw was that, for streams, the views dont change too much after they are a 2 or 3 days old, for example, after those few days, a 200k views would grow to like 210k. The older they were, the less % they grew. If you want example in numbers, i can get you any vtuber number collected by the weekend end, you can compare with what you see now (see below). Either way, every streamer is subject to that same effect, so it evens out as all of them should grow a bit.
BUT, thats for streams, the type of videos i counted. Songs and short skits are a totally different matter. I remember specifically Towa and Mori's original songs: they still kept a steady and almost constant growth, as you said, between those two weeks i checked for their streams. If you want to compare songs, you might have to consider the time after release, as you said, indeed.
here's an example of stream vod views collected 3 days ago. I dont have the tiles, but you can match them by their duration times:
Ina: https://www.youtube.com/channel/UCMwGHR0BTZuLsmjY_NT5Pwg/videos
3:14 225
2:24 177
2:54 127
3:42 219
2:59 214
3:11 200
3:02 132
2:04 179
2:05 149
3:27 179
2:02 170
6:15 436
2:32 187
1:19 93
3:07 174
3:27 186
2:30 135
1:34 122
1:50 116
1:55 194
I cant give you the full short songs examples working on different behavior though, because i didnt use them for my tables. The one I can remember best: is Matsuri&Gura's Soraioro Days cover: it reached 1mil views in 1 to 2 weeks, and now on 3rd week its still growing a lot, being at almost 1,7mil atm.

[Anonymous No.]

>>2748133
Oh no problem, I wasn't talking about your table (because sure, if you only take rather recent videos into account, the "accumulation over time" thing isnt not yet that important and doesn't really need to be so much adjusted for), rather I was talking more in generalities about an overall point when doing view analysis (ie. that an X amount of views reached in a shorter time is "better" than the same X amount of views but reached over a longer time).

Keep crunching them digits, fellow numberfriend!

[Anonymous No.]

>>2748133
Maybe I missed something but isn't the same effect at play in stream length? The longer it goes on the more the effect of the length of the stream should diminish as at some point everyone who is going to watch it will have already been watching it.

[Anonymous No.]

these threads is just skewed stats just to make holo look good. Why do you need constant validation holy shit. Streaming numbers don't count unless it's a holo and they're winning at it? Fucking stupid

[Anonymous No.]

>>2748355
yes, you can look at the ratios on the table, for every single streamer, there is an overall trend that, the longer their stream durations get, the lower their ratio of views/time get as well, because they go past their prime hours into dead hours, and there is less renovation in general.
Thats why I only compared the ratios of same time hour lengths, so that effect gets greatly diminished.
To give an example, viewes/time 2.12k ranks 1st for the 6h<time<7h time length, but if that 2.12k ratio was for a 1h<time<2h, it would only rank 24th.

[Anonymous No.]

>>2742848
Ame does tend to stream a lot with superchats turned off so her numbers don't really surprise me.

[Anonymous No.]

>>2748655
it's a form of holocope kek

[Anonymous No.]

>>2750785
>it's a form of holocope
>it shows HoloEN has one the biggest falls