[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/vt/ - Virtual Youtubers

View post   
  [?]In ghost mode, only threads with non-archived posts will be shown
File: 82 KB, 1080x392, Screenshot_20230507-122748.jpg [View same] [iqdb] [saucenao] [google]
48893947 No.48893947 [Reply] [Original]

Why is there such a discrepancy in live viewers and votes? Is youtubes viewer count wrong?

>> No.48894061

>>48893947
it's literally designed to be wrong on purpose

>> No.48894098

>>48893947
Because Youtube is a piece of shit masquerading as a video sharing website written with code so fucked up people could mistake it for the gordian knot.

>> No.48894240

>>48893947
People join and leave during the polling. It's rare for people to stay longer than a few minutes in any stream.

>> No.48894282

>>48893947
Because of viewer turnover
most viewers do not stay for the whole stream

>> No.48894378

>>48893947
EN viewers just fucks off and nit finish the stream

>> No.48894411

>>48893947
>enter stream
>vote
>leave stream after 10 minutes

>> No.48894519

>>48894098
This. Despite being one of the biggest online platforms, the entire site is shite. Half my screen doesn't work sometimes on vids, and I know it's not me since it works in any other app.

>> No.48894559

>>48893947
because it's Minecraft. unlike faggot in jp, westoid prefer more engaging games

>> No.48894635

>>48893947
Because if you have adblock you don't count as a human for youtube

>> No.48894646

>>48893947
Because people with ad-blockers aren't included in the CCVs and view count of videos and streams. Considering Ame's history, uniqueness, and fanbase. I'd imagine a lot of her viewers use adblock.

Polls are actually a fairly good judge of how many people are actually watching, since about only 40-50% of the people watching a given stream will vote. The others are just listening while doing something else.

>> No.48894655

>>48893947
If I'm not mistaken, people with adblockers aren't included in the live viewer count, nor are people that are tabbed out/don't have the stream as the focused window. There's probably some other stuff gets people excluded

>> No.48894705

>>48893947
Youtube doesn't count you unless you're using chrome with no extensions whatsoever and are tabbed in or are using youtube on android.

>> No.48894951

wait so this whole time EN who has viewers for like 2k or 6k actually has 10k all along ? . . youtube fucker

>> No.48895185

>>48894646
>Considering Ame's history, uniqueness, and fanbase I imagine a lot of her viewers use adblock

Curious at what you mean because adblock and those things have nothing to do with each other.

>> No.48895356

>>48894240
>People with engagement so low they can't stay more than a few minutes take the time to vote on a poll they don't understand or care about.

>> No.48895543

>>48894951
Twitch CCVs are always considerably higher than youtube CCV for this reason

>> No.48895546

>>48895185
Ame's fans all have cuckold fetishes and people with cuckold fetishes use adblock because they frequent porn sites that have a lot of ads

>> No.48895559

>>48894646
So going off of this, ame has like 32k viewers rn. How many people were actually watching the recent massive tetris stream then? Millions?

>> No.48895593

>>48894240
I can confirm. My normalfag senses go off whenever a poll is made and tune in to a stream to vote then immediately leave.

>> No.48895637

>>48894951
yeah, I am pretty sure there have been polls gura put up where it says she had 16k viewers but the poll had like 40k

>> No.48895681

>>48893947
obviously bot voting

>> No.48896116

>>48895593
Kek, keep up the good fight anon

>> No.48896189

>>48894411
You forgot the
>refuses to elaborate

>> No.48896253

>>48895356
>People with engagement so low they can't stay more than a few minutes take the time to vote
seconds. not minutes. seconds. the churn required to justify youtube live viewer counts is so fast that people would have to literally open the tab, vote, and instantly close it.

>> No.48896410

>>48895681
Do you have the pic of Sara when they botted, I was in that stream but was playing only in the background

>> No.48896455

>>48893947
the real answer is that youtube culls views extremely hard. if you don't have the stream at 480p, above half volume, focused for more than 30 seconds then you don't count as a view.

>> No.48896540

>>48895356
>Retards click random buttons shoved in their face without understanding them
>This is surprising
You can look at what average analytics look like if you want. Most people do not click on a stream to watch for two hours. They click to kill time for five minutes.

>> No.48896557

>>48894240
The fact people still don't realize this is mind boggling. 90% of this thread are retards going off about ad blockers and stupid conspiracy theories. They are so dumb they don't realize people LEAVE streams. I don't usually feel smarter than everyone in the room, but you dumb cunts are retards

>> No.48896920

>>48895356
You’re retarded, literally everyone does this, you probably do it too as a matter of fact

>> No.48897096

>>48893947
Most people don't watch for more than 30 seconds, and YouTube will only count them if they stay for longer than that. It also counts as an interaction (just like liking and commenting) which encourages the algorithm to recommend it to people. This was tested on Spiffing Brit's live stream where he told his users to refresh all at once and 30 seconds later the time graph of impressions spiked (also yes YouTube considers refreshing a video/stream as a new view).

>> No.48897291

>>48897096
suggesting with a straight face that thousands and thousands of people, within the span of one minute, open a stream, vote in a poll, and then leave, is absolutely insane. and i've seen multiple polls that exceeded the viewer count so quickly and by so much that the only possible explanations are that thousands and thousands of people did that, or that youtube is grossly underreporting live viewers. youtube is grossly underreporting live viewers and it is a hard fact.

>> No.48897301

>>48897096
Have you ever tried the tea he has shilled? Thinking about getting one

>> No.48897533

no one watches live streams, everyone uses youtube, simple as.

>> No.48897776

>>48897291
I imagine that people who like clicking on random videos without even watching them are also prone to clicking random polls without even reading them. They're probably mostly children, at least mentally.

>> No.48897868

>>48897776
i do not believe that thousands and thousands of people do that to hololive live streams within the span of less than 60 seconds. that is literally insane.

>> No.48897930

>>48897776
It's way too fast. You have to be there.

>> No.48898314

Vtuber who has a subway surfer video on the right half when.

>> No.48898773

>>48897868
It's funny, this would make them fucking gold for sponsors in terms of exposure. It means you're feeding your advertised game or product to an insane number of viewers who voluntarily tune in to look at your ad without the con that they would stay and get spoiled/disillusioned when the rough edges inevitably show themselves.

>> No.48899111

>>48897868
There are a lot of 10-13 year olds on YouTube despite the TOS explicitly forbidding them from using the site.

>>48898773
What good would that do when they aren't even paying attention?

>> No.48899342

>>48899111
>There are a lot of 10-13 year olds on YouTube despite the TOS explicitly forbidding them from using the site.
no, they don't open, then click a poll, and then immediately leave. it's a fact that it's grossly undercounted. i have watched this live. i am not using statistics sites later. i have watched with a count of under 20k, then seen a vote of 30k fill up within under 60 seconds, and observed as the entire time the stream never reported over 20k. it's grossly undercounted, and it is a hard fact.

>> No.48899493

>>48898773
not all impressions are treated equally u fucking drooling idiot, look at esports just as an example, lol no one even knows who these girls are and their twitter reach is dogshit for how big their accounts are, you have no idea what you are talking about and what companies actually look for in 2023 when it comes to sponsors.

>> No.48899546

>>48899342
Those underage kids are quick little devils.

>> No.48899620

>>48894240
But that makes zero sense. If people left the stream after a while why would the votes be higher?

>> No.48899870

Does add blocker users seriously don't count? Can they still type in chat?
I'm actually intrigued by this, i'll go do some research since i have nothing better to do.

>> No.48900091

>>48896557
because /vt/ have no live, they tought everyone is like them who watch vtubers for hours every day.
people have things to do, some of them stay for a while and leave for other things.

>> No.48900129

>>48894240
Very rarely something like this happens on other streams, and when it does, is very likely to be another hololive stream.

>> No.48900328

>>48893947
youtube is really picky with what it considers a real time viewer
Which is different from twitch where i could have 3 streams on 3 differnt tabs, all muted and still count as a viewer on the 3 of them.

>> No.48900527

>>48900129
It's pretty normal on YouTube for people to click through recommendations willy-nilly. Maybe you're thinking of Twitch where changing streams runs the risk of getting a 30 second preroll ad.

>> No.48900880

>>48899620
Fucking retard. Likely nigger.

>> No.48901296

>>48900328
someone who is watching during the minute that a poll is posted and then votes in the poll is a viewer, full stop.

>> No.48902668

>>48893947
I've literally watched a vtuber with several people chatting (including myself) and Youtube said she had zero viewers.

Numbers will and always be garbage; streaming sites have every reason to lie about the numbers.

>> No.48902844

Wasn't it Coco who mentioned the average time watched per viewer was something like 4 seconds? Or was it someone else? Which is why they're usually more relaxed in membership streams. It's not as much about "I trust you guys, we're so close hehe", it's more that they don't have to worry about making a good first impression every 5 seconds.

>> No.48903004

>>48893947

She had it running the entire stream. People join, people leave. It's not a hard concept.

>>48895356
Literally "what is best melon" with 2 choices; such a hard question for people to understand.

>> No.48903038

>>48893947
Because your used to twitch. On Youtube you actually have to be viewing the page.

On twitch you can have 10 tabs open and you count for all 10 of them even if you have them all muted. Thats why Twitch engagement is always shit like 30%. If you have 10 tabs open in youtube only the one your watching counts. What you have it muted? Yep your view doesnt count either.

>> No.48904189

>>48903038
>Twitch's engagement is 30%
>Youtube's is 300%
Which one is a better?

>> No.48904409

>>48893947
there was that one time one of those sony vtubers had like 80k ccv in minecraft and less than 200 votes in a poll asking if they were real

>> No.48904813

>>48901296
not to youtube. there is a bunch of never explained rules that can influence if you count as a viewer or not. you can test some of it by making a stream and having a friend check things like watching it with or without addblock

>> No.48904851

>>48893947
YT is only counting live viewers that it can "identify". On the other hand, everybody who spent few seconds on the stream can vote even if they are going to leave soon.

>> No.48905074

>>48894240
I left when I realized cantaloupe was actually losing. Couldn't stand sharing the stream with honeydewfags.

>> No.48905076

>>48899342
I can guarantee you have never seen 20k CCV turn into 30k votes within a span of 60 seconds. The vote count is ALWAYS less than the CCV. The only way for the vote count to be higher than the CCV is when the poll stays up for an extended period, allowing new viewers to vote on the poll as they replace old viewers who leave. I have never seen CCV exceed votes within 60 seconds. It was probably 3000 and you misread it as 30000.

>> No.48905217

>>48894240
how would they even fucking know there was a poll if they don't have the stream open? Yeah that's what i thought fucktard

>> No.48905233

>>48904813
>not to youtube
i utterly reject this interpretation.
>>48905076
>I can guarantee you have never seen 20k CCV turn into 30k votes within a span of 60 seconds.
i'm an ex-chumbud.

>> No.48905587

While I love shitposting about girls dying because "X" as the next guy.
It's been obvious since like 2021 that youtube stopped counting most people as actual viewers. We already had this check before by Nene, Calli and Kiara where they did rapid votes so new people couldn't "magically" show up and they each ended up with double the votes to the amount of viewers they had.

>> No.48908204

no idea if you need a youtube account to vote. maybe a bajillion people watch without account and can't

>> No.48908272

>>48893947
I wanna say adblockers remove the numbercount, which already takes weird as fuck snapshots

>> No.48909456

>>48894098
>>48894519
The fact that there are around 12-15 different teams, with barely any communication to each other, working on the site should be enough to tell you that YT is the world's biggest case of spaghetti code.

>> No.48909712

>>48893947
yes, YT numbers are basically made up
/#/ is really dumb

>> No.48909733

>>48894061
Explain.

>> No.48909799

>>48896557
there have been holo polls that have gotten three times the CCV in votes in two minutes
thousands of people all just magically tune in when a poll is posted, quickly vote before the system sees them and leave?
you are a fucking retard

>> No.48912265

>>48896557
That doesn't change the argument at all, in fact it makes polls even more valuable because what they mean is the amount of people who clicked the link to the stream at all.

>> No.48912392
File: 6 KB, 171x49, 5min poll 10k viewers.png [View same] [iqdb] [saucenao] [google]
48912392

>>48894240

>> No.48912572

>>48894559
You know anon having ADHD don't prevent you from being a faggot, actually quite the opposite so i'll repharse your error : unlike chad and healthy JP, ADHD amerifag can't keep their attention on a single stream for more than 10 minutes before having to go shitting due to their terrible diet that make them fatsos.

>> No.48912634

youtube doesn't count anyone using adblockers as a viewer. Many people use adblockers. Simple as

>> No.48912639

ITT: newfag finds out about youtube cull numbers

>> No.48912686

>>48894240
yeah that's what normal people do, they join, vote on a poll they have no investment in, then leave

>> No.48912728
File: 2.77 MB, 1456x470, they simply lost interest 5 seconds into the stream.webm [View same] [iqdb] [saucenao] [google]
48912728

>> No.48912788

>>48902844
I mean when Magni talked about it he said the average viewer watched for 5 minutes, left a like, then watched for 5 minutes more before leaving.

>> No.48912885

>>48912634
Apparently it's the "||youtube.com/get_video?" entry in the "EasyPrivacy" list. Turning it off doesn't re-enable ads either, must just be some "I don't want -anyone- to know what youtube videos I'm watching" super autism option.

>> No.48913075

>>48893947
Either it's h1b incompetence encouraged by troon managers or it's intentional in that they siphon off viewcount and "like" stats and distribute them to vods and streams of bullshit their masters tell them to promote, whether they're being paid to or they do it for free because of their ideology. Either way they defraud their advertising customers this way.
Even if they're not actually just "stealing numbers" they can simply dampen things they don't want to get popular so that they don't seem as good vs things they want the algorithm to push "for reasons" (ESG, personal politics/religion, corporate bribes, government orders).

>> No.48913753

>>48894240
I'd say that the answer is that it's both YT culling based on various factors (embedded videos don't count so Holodex immediately comes to mind, users with adblockers, users who don't have the stream on an active tab, even if the stream itself is still on top and shown to the user, users with non-Chromium browsers, users from certain regions, etc.), and viewers arriving and leaving during the stream. For a 6.5k stream, whenever the CCV counter refreshes, it shows a delta of ~20-30 viewers in every interval based on previous experiences, so counting for both viewers who join late and viewers who leave, we have roughly a 0.5% turnover rate for every refresh of the counter. Now I don't know the refresh rate, but I'd say it's about 15 seconds, which means that on average, your viewerbase who can realistically vote (i.e. watched your stream for at least 30 seconds or so) is twice of your CCV by the time your stream reaches a 100 minutes. Now this is some rough napkin math and obviously this doesn't mean that all your viewers shift over over the course of 100 minutes, it's much more likely that those who tume into late are not invested and leave after voting and watching for 10 minutes or so, but the turnover viewers in polls are still significant.

>> No.48914085
File: 46 KB, 1024x660, adblockers.png [View same] [iqdb] [saucenao] [google]
48914085

>>48893947
YouTube team usually "explains" this by stating that likes and votes do not go through the same filtering process that viewers go. They hope you just accept that those are not accurate and move on. There are a few things you need to consider before addressing the elephant in the room:
1.Some streams clearly have botted activity, but assuming that all of them do that would be borderline ludicrous
2.Building on #1 point, YT has extremely high live viewers engagement (voting). This means that if you make a poll that expires in 5 minutes, it would not be uncommon to get 60-80% engagement, but that is not in line with other streaming platforms where viewers engagement would rarely exceed 50%. And no, streams don't have such high viewers rotation to make those polling numbers inflate like a balloon in under 10 minutes. Usually we'd be looking at less than 5% error margin due to viewers tuning in and out, but not 20-30%
Now to address the actual problem that's causing this: if you have ad blocker on, or you are tabbed out, you will not be counted toward CCV even if you're logged in. YT treats everyone with ad blockers like trash not worth counting. They most certainly know YOU are watching because you're using their chat while being logged in, subscribed/membered or even sending SCs, but they still choose not to acknowledge your presence (with a few exceptions). Why? Your guess is as good as mine, but I'd wager it's because YT is still primarily a VOD platform that wants to retain their high $/1k views payout, but you can't have that if half of those views are unmonetized.
Unfortunately for streamers, the number of ad block users will only go up (see picrel), and as long as streaming plays a second fiddle to VODs on YT, that shit will never get changed. Yes, it doesn't make sense considering how they should be aiming to show as high CCV number as possible, especially to attract more streamers, but then again, is this really surprising considering all of the retarded decisions they've made over the past 10 years?

>> No.48914213

>>48912728
me when I click a stream and realize it hasn't started yet so it's going to be 17 minutes of a streamer's generic "Starting Soon!" gif with a 40 second audio clip on loop

>> No.48914221

>>48914085
Yeah bilibili has like 1% engagement obviously something is wrong with youtube.

>> No.48914366
File: 59 KB, 719x707, 1621924840174.jpg [View same] [iqdb] [saucenao] [google]
48914366

I can't believe there are retards questioning this. Do you also wonder how a stream that had 5k ccv manages to have 40k total views after they finish streaming for an hour?

>> No.48915572

>>48895546
every person with a functioning brain uses an adblocker when browsing the web

>> No.48916084

be honest, how many of you actually bother to vote in polls?

>> No.48916880

>>48916084
I only vote if I have a strong opinion. So like, for the Ame poll posted earlier, no, I did not vote.

>> No.48916999

>>48896557
You missed 2 separate fiascos where youtube clamped down even harder on "fake views" and suddenly some holomembers were 3 views and outright did polls to see how many people were actually watching. They instantly had 5 digit vote numbers on the polls in minutes.
When people contacted youtube support about it they said nothing was wrong for a week until their twitter got shitspammed with posts about it and they finally fixed it.

>> No.48919165

>>48894061
this

>> No.48920141

>>48893947
>click on stream
>sees poll
>vote
>5 mins have passed
>close stream

>> No.48921203
File: 1.43 MB, 1522x785, baepoll.png [View same] [iqdb] [saucenao] [google]
48921203

>> No.48924314

>>48920141
This

>>
Reply to thread [?]Don't expect anything heroic. Your post will not be uploaded to the original board.
Name
E-mail
Subject
Comment
Action