White House YouTube Dislike Manipulation | Hacker News Reader

White House YouTube Dislike Manipulation

phzoe.com

65 pointsauganov37 comments

YT also removes millions of likes/views, for example on K-pop videos which are fake-streamed by their "armies" of fans.

There are tutorials on the net on how to "ethically", "honestly" increase view counts on your favourite band, with stuff like no more than 1 watch every 20 minutes:

https://twitter.com/bstwings_views/status/129646245602223718...

So it's not easy to tell what's going on here, they could be inauthentic dislikes.

The article implies something nefarious going on here, but there are plenty of entirely rational innocent explanations. i.e. the videos are attracting a large quantity of spam dislikes, and YouTube is bulk-removing them.

> a large quantity of spam dislikes

what exactly is a 'spam dislike' . I don't see how thats a valid explanation.

Probably bots, people who do nothing but dislike videos, Indian sweatshops that are human bot farms, dislikes driven by social media posts calling for users to add dislikes, etc.

> people who do nothing but dislike videos

> dislikes driven by social media posts calling for users to add dislikes

this is not spam. how does youtube knows which dislikes are driven by 'social media post' ?

> Indian sweatshops that are human bot farms

someone is paying sweatshops to dislike whitehouse youtube? How do you know this ?

>> people who do nothing but dislike videos

>> dislikes driven by social media posts calling for users to add dislikes

> this is not spam. Also how you think youtube knows if a certain user is disliking because they read a social media post ?

I believe Reddit calls similar activity "brigading," and it's against the rules there.

>this is not spam.

Dislikes are there to help human users of Youtube (and the AI recommender systems of Google) not to be some sort of god given metric of anything. Spam is anything which makes dislikes less useful as metric for those things. Users who dislike everything aren't a useful metric.

>Also how you think youtube knows which dislikes are driven by 'social media post' ?

A sudden jump in dislikes with a referal tag of <insert social media>. Probably mixed with their knowledge of those user's other Youtube browsing behaviors. And Google's absurd knowledge of your internet browsing behaviors via its massive analytics network. I'm sure there's whole teams doing nothing but this at Youtube and I suspect they have very sophisticated methods at this point.

> dislikes driven by social media posts

(Disclosure: I live outside the USA, I'm not politically active, I have no horse in this race)

It's clear from the way that many people write that if "their side" is doing something it's almost certainly the best of reasons and if "the other side" is doing something it's for the worst of reasons.

From the HN guidelines: "Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."

How can we reintroduce good faith into everyone's interactions with the political class? Right now it seems it's perfectly OK to assume the worst possible interpretation of anything a politician says*

When did the world get like this? How do we fix it?

* if s/he isn't from your party. If s/he is from your party, of course you assume the best interpretation.

This is nothing new in terms of US politics and political candidates gladly fan the flames:

https://www.brookings.edu/blog/fixgov/2019/07/11/a-short-his...

edit: And if I remember in Rome a politician was literally killed by a mob of his opponent's constituents so this has been going on for thousands of years.

One way I have seen this done in the past is that you have contest sites. As a user you gain x number of entries for every action you take. One of these actions can be to go in and like or dislike a video for example. Sometimes the sites can even check that the action was performed.

Someone wanting more entries might create lots of fake accounts.

I'm sure there are many other ways to do this but that is one example on how spam dislikes can be generated.

But what would someone get out of paying ppl to dislike whitehouse videos ?

> But what would someone get out of paying ppl to dislike whitehouse videos ?

Propaganda.

But how do you know this is happening for this video though?

OPs original assertion was that this has ' innocent explanations.'

Indian sweatshops, contest sites to spread political propaganda ect are conspiracy theories with no proof, far from 'innocent explanations'

>>> But what would someone get out of paying ppl to dislike whitehouse videos ?

>> Propaganda.

> But how do you know this is happening for this video though?

It's still the answer to the question you asked, whether it's actually happening or not.

Ah yes you are right. You were responding to my question.

I had all the replies mixed up.

> the videos are attracting a large quantity of spam dislikes

Prior to the bulk-removal the dislikes are tracking the views; the dislike growth curve is a close function of the view count curve. That is unlikely to happen if the dislikes are bulk spam.

We'll be told it's 'spam' or 'bots', and Youtube will rely on this as the plausible explanation should they bother to address it at all, but the inconvenient truth is that the most likely explanation for the dislikes is viewers. Otherwise one must posit a sophisticated spammer that is carefully tailoring their dislikes to camouflage their activity.

I suspect the only meaningful response we'll get is a throttling of the API used to do the analysis.

Other comments here have pointed out that the drop is likely due to a periodic cleaning of bot votes, against the blog author's premature, sarcastic insinuation that some nefarious manipulation is being performed at YT (there is, but it's not nefarious but to combat third-party manipulation that misleads the user and abuses the platform).

Perhaps it is also of note that the author performs a naive analysis of limited climate data in a linked post to draw a conclusion against climate change.

There's still like 10x the dislikes then the likes. So a pretty crappy conspiracy if it doesn't even get a positive like to dislike ratio. Or it's a whole bunch of bots/spam mixed in with legitimate dislikes.

For a while YouTube didn’t report the exact number of views, likes and dislikes. So someone viewing a video would only getting an approximate value for each or with a deliberate small +/- offset. Is it possible that’s what’s going on here?

I find it hard to believe YT is engaging in “manipulation” of a stat that frankly doesn’t really matter if you’re an account like the White House.

>> of a stat that frankly doesn’t really matter if you’re an account like the White House.

I don't agree. I think these numbers matter more than we realize. I believe that many politicians today are using Twitter/Youtube/Facebook likes as modern Gallup polls. Manipulation of the numbers could result in actual policy changes.

Say there is a video about net neutrality, or the breakup of the social media giants. I honestly think all sorts of people in government will be looking to the popularly, or lack thereof, as a gauge of public support for the content of the video. So, imho, manipulation of likes/dislikes is becoming very serious.

I think you are mistaken if you believe anyone in government cares about number of likes and dislikes on any YouTube videos and uses this to base their position on any given policy.

Even if they cared about traditional old-school polls you’d probably have Medicare for All, the USA would be completely out of Iraq and Afghanistan and you’d be able to buy Cannabis legally. I doubt they’d ignore the known popularity of these things but suddenly care about social media likes and dislikes.

Totally not dislike bots.

Assuming sarcasm, that was my first thought as well. Although it's possible there is something nefarious going on, I assume the White House attracts a disproportionate number of bots and this could potentially be accounted for by purge of dislikes by bots.

As bad as it looks, it could be an automated attempt to protect the system from manipulation. As such, I find the author's conspiratorial tone somewhat unjustified. (For example, the comment about 1 am probably has little to do with doing it when people are unlikely to notice it and more to do when maintenance is done.)

Is it possible a "DisLike" botnet would follow a similar pattern? Find a video, dislike for a while, find the next, repeat.

It seems obvious there would be plenty of "legitimate" dislikes coming, as probably a lot of people who followed this YouTube channel in the past 4 years were Trump supporters, and now the channel features Biden. Trump supporters still subscribed would dislike this content. I wonder if the same happened at the last transition.

> Trump supporters still subscribed would dislike this content. I wonder if the same happened at the last transition.

I'd be surprised if it hadn't. The bigger question is whether YouTube manipulated the dislikes back then, too.

> Trump supporters still subscribed would dislike this content. I wonder if the same happened at the last transition.

Presumably you could unsubscribe. Which is not the case with instagram. If you followed the White House account, or you decide to follow the archived account for Trump, you are periodically made to follow the current White House account for Biden. Even if you block the user, it will unblock and re-follow you.

> It seems obvious there would be plenty of "legitimate" dislikes coming, as probably a lot of people who followed this YouTube channel in the past 4 years were Trump supporters, and now the channel features Biden. Trump supporters still subscribed would dislike this content. I wonder if the same happened at the last transition.

I don't know about YouTube, but I've read that the 2017 transition and the 2021 transitions were different on Twitter:

2017: Obama's institutional twitter accounts (e.g. @POTUS) were archived, but Trump inherited the followers in his new @POTUS account.

2021: Trump's @POTUS was archived, but Biden's new @POTUS didn't inherit Trump's (and Obama's) followers. He had to start from scratch.

https://www.nytimes.com/2021/01/19/technology/biden-white-ho...

The videos all seem to have far more dislikes than likes (very different from the article).

They actually seem to be quite aggressively disliked, disproportionately from standard polling anyway. Maybe there is something nefarious happening, but not at YouTube.

This website also contains climate change denial and scare quotes around the word scientist.

We should be more skeptical of sites like this.

This is whataboutism and trying to discredit the researcher. Either analyse and critique the very concrete work displayed here, or don't say anything at all.

Examples?

https://phzoe.com/2020/03/04/dumbest-math-theory-ever/

https://phzoe.com/2020/10/05/geothermal-to-the-moon/

https://phzoe.com/2020/02/20/two-theories-one-ideological-ot...

There are may others. It's all over the place. The author denies the Stefan–Boltzmann law that contradicts her model meant to prove most heating is due to geothermal. She also does not believe in Fusion within the Sun, or that photon are particles (only waves, which is linked to the thermal radiation denial).

Additionally, the author mocks and attacks anyone contradicting her in the comments. Her argumentation revolves around throwing new theories after new theories that could help explain her impossible results, expecting others to provide the contradiction.

All-around extremely exhausting to argue with, bad faith, aggressive and ideologically driven.

Amidst all this, many election fraud allegations on the blog, beyond the subtext implying manipulation in this post article.

Thanks and fair enough.

The person also thinks the GA runoffs were fraudulent, for the record.

Has this been observed before on any other videos?

Reply on news.ycombinator.com