Hacker News new | past | comments | ask | show | jobs | submit login
YouTube suspends Google accounts of Markiplier's viewers for minor emote spam (reddit.com)
153 points by bluefreeze 26 days ago | hide | past | web | favorite | 95 comments

The real issue here is that (reportedly) the people who spammed got their GMail accounts shut down.

That's outrageous.

That's your real-life, not just your online-life. Your GMail account isn't even related at all to your YouTube identity.

Remember: if you aren't _at least_ paying for your e-mail service you don't own your own mailbox. When was the last time you made a backup of your GMail account? What contingency plan do you have in-place if Google ever accidentally, unintentionally, or in this case, intentionally, shuts it down?

(Disclaimer: I pay $20/mo for an Office 365 E3 account for my personal mailbox)

Google should be able to disable access to a specific service, not shutting down the whole account.

They want to be the identity of the online world, they need the responsibilities that comes with it.

When I was a kid I had a Yahoo site with some images hotlinked to another domain. The domain owner changed the images to porn, reported my Yahoo account, and it got my account entirely deleted. Emails and groups and chat and all. Really dumb

Since when has Google ever took on "the responsibilities that come with it"?

I remember a story (maybe from here?) a couple years back where someone had their gmail account blocked - and they lost google services - and they needed access to docs and other things to work with his team.. and after trying for days to get it brought back, he was essentially laid off from work as he could not participate in any of the tools that were being used by the rest of the company.

I had a time a while back when my youtube account got blocked - and then I found that googleplus had put a block on the account and it auto-blocked the youtube videos - and they said all the videos were find for youtube and no problem there - but they would not go over the heads of the googleplus team, so we were screwed.

we should have posted on many more sites your advice about "you don't own your own mailbox. When was the last time you made a backup of your GMail account? What contingency plan do you have in-place if Google ever accidentally, unintentionally, or in this case, intentionally, shuts it down?"

like how this site has recently added info about big G's censorship: https://news.ycombinator.com/item?id=21490272

most people don't know about these issues - they just assume gmail will be there, and google's search shows truth.

Or self-host. Linode VPS: $5/mo and for that I can host E-mail, a light duty web server, and other various odds and ends. I don't have to worry about some random algorithmic account suspension removing access to my E-mail (and all of my online accounts, since they use my E-mail address for verification). And I don't have to worry about some random algorithmic process determining what E-mails I receive and which ones get silently dropped into a black hole. So far no problems with misconfiguration or other providers dropping my mail--while it probably happens, I think the risk of that is overblown if you configure SPF, DKIM, and so on correctly.

We're there any good resources you used for learning how to configure SPF and DKIM?

Linode's documentation tends to be pretty good: https://www.linode.com/docs/email/postfix/configure-spf-and-...

That guide assumes you're using postfix as your MTA. For exim, their documentation can be found here: https://www.exim.org/exim-html-current/doc/html/spec_html/ch...

Distributions often have their own documentation, for Debian: https://debian-administration.org/article/718/DKIM-signing_o... and https://debian-administration.org/article/719/Avoiding_mail-...

Keep in mind this is all to increase deliverability of your outgoing mail. You may also optionally want to check SPF and DKIM on incoming mail to cut down on spam: https://debian-administration.org/article/721/Validating_SPF...

We need legislation. Google is too powerful.

Legislation to protect e-mail mailboxes as the same way we do with official postal (snail-)mail services in most countries needs to be handled very carefully, because otherwise we'll end up protecting spammers - the very opposite of the people we want to protect.

I see this story as Google's equivalent of when Valve would ban people from Steam for crude comments they made on the Steam forums, which I understand meant they forfeited their entire game library - or Amazon banning people for reviews.

If there is going to be any law or legislation - it should be to require service companies like Google, Valve and Amazon to treat the customer's "online personality account" (i.e. forum account, "community" account, etc) separately from their "value" account (i.e. whatever that which has _value_, be it their Steam library, their GMail mailbox, etc) - or at the very least, companies must be required to export 100% of the data in an account and providing that to the customer/user in the event they shut down an account (with reasonable constraints, such as the account having existed for at least 3 months and having been in good-standing until this specific incident, etc).

And under no circumstances should anyone lose their software copyright licenses for violating a community ToS - feel free to ban them from the forums, et cetera, but being banned from the Steam forums must not mean they lose the right to play the copy of Half-Life 2 that they paid for.

Nit: Valve just banned them from forums, not Steam. They already do what you're proposing.

Got a link? I remember reading that people were banned from their Steam accounts entirely without any refunds and they had to resort to CC chargebacks.

No link, but I got a personal anecdote.

I found a persistent XSS vulnerability on steamcommunity.com once upon a time. It used to be if you were playing a mod, the name of the mod could be shown in activity graphs and screenshot views. This was not sanitized whatsoever.

After 4 weeks of no response from their official email contact, I posted it on the (then entirely detached vBulletin based) Steam forums, with a notice I had disclosed the issue to all known to me @valvesoftware.com email addresses (including security@, which didn't bounce)

I had my account disabled, and spent 3 weeks emailing them asking what the fuck that was all about. It got re-enabled silently.

Well, fairly simple to solve; mail services should be able to restrict sending mail (possibly only to family contact, limit of 25 contacts or so) until the matter is handled, ie temporarily if it's not blatant abuse of the service (ie spam). Otherwise, receiving mail should be a right once you have a mail address at a larger provider.

Why isn't "use a different email provider" an option? Why do we need laws to enforce good customer service?

For a large part of the population, losing their email account could be devastating. "Could" is deliberate, because once you lose it you wouldn't even know whether anything important landed in your inbox. Like a personal update from a friend, or some official business you need to respond to within days. Can you imagine trying to remember every single friend, business and government agency which might conceivably contact you, and figuring out how to contact each of them to let them know of your new email, plus getting them to re-send anything since you were blocked? It could put a wrench into someone's life for days or weeks.

I understand the problems losing access to your email can cause, in fact, I have experienced this exact problem several years ago. It took me 3 weeks to recover my email, and after that experience I switched from gmail to protonmail.

Most people won’t switch until something like this happens to them. We shouldn’t go, “just switch!” because it won’t work. And we definitely shouldn’t go, “screw you if you didn’t switch because you didn’t listen”

Yeah, the reason most people don't switch is because it doesn't happen to most people, ever. If you don't like a company's service you switch to a different one, that's how capitalism works, especially services you pay nothing to use.

More importantly, pay for your own email DOMAIN.

And keep backups of your email.

Then, even if your account gets shut down, you can get up and running elsewhere quickly.

Be careful:

By hosting your main email on a custom domain(not a provider) you open a new attack vector for identity theft.

Someone can hijack your godaddy/namecheap/gandi account and point the MX DNS records of the domain to their own server and receive all your "Forgot your password? here is the link to reset!" emails

This a very bad advice unless you actually know what are the risks.

Is there a good way to protect against this aside from good password hygiene at your domain registrar?

I use namecheap if that is relevant--maybe there's a better/more secure registrar to move to.

But I guess, equivalently, someone can just equivalently hijack your email account directly if you use a service like Gmail or Yahoo.

I think namecheap does not have a bad reputation, godaddy is known to be bad though(both security and customer support).

The thing is, of course people can hack your gmail or yahoo mail somehow, but when you host your own domain, you have the same attack vectors + dns hijacking and this last addition is easier to exploit with social engineering. Google atleast is virtually immune to DNS hijacking.

Just the usual advice, I would think: Strong, unique password, and enable 2FA.

It may helps if the sender send you password-reset email in encrypted format

How is that any different than someone hijacking your mail account directly? I mean, yes, now there are 2 things to secure, but it doesn't seem like very much added attack surface.

No only that but, by doing so, they're effectively taking away any remaining semblance of proof that you weren't spamming.

Worth noting: I'm not sure about Microsoft's procedures for shutting down an account the way Google did here, but they document what happens if you stop paying for it: https://docs.microsoft.com/en-us/office365/admin/subscriptio...

Basically it goes through Expired, Disabled, Deprovisioned. During Expired users have access, Disabled only Admins have access, Deprovisioned is gone. Those are usually 30/90 days but it varies by how you purchased.

Not sure about Personal/Home, but I suspect it's similar.

Yeah.. True that. I lost access to paid email account due to problem coming up with funds for paying this year's fee. Was not a fun 2-3 weeks.

I am with you but I think protocols and apps that are designed to allow users to own and host content on primarily their own device would be great.

For mail, what if email was always delivered to device for free and you pay a service provider to mirror it so that while your account is active,they provide backup,spam filtering and content search on their servers?

I would also recommend never storing mission-critical stuff only in your email. Losing the contents of your inbox should be the least of your concerns if there’s a chance you might lose access to your account. There are far worse repercussions - e.g. a) how do you update everyone else's record of your email address? b) how many services have you locked yourself out of, or made doing so much more likely?

It makes owning your email address (including the domain name part) crucial.

It's time people separated the hosting of email from the interface (the address).

I think most people just don't realize how easy this is and you can even keep using gmail.

Most domain registrars have free email forwarding so you can forward *@domain.com to your gmail. This by itself is already good enough for signing up for things or resetting passwords.

If you also need to send from that domain in gmail there's just a few more steps: https://support.google.com/domains/answer/9437157?hl=en

Unfortunately it is no longer free to use a custom address in gmail; it used to be that just owning a domain would do the trick.

I wish lawmakers instead of spending time on misguided cookie banner laws, spent more time on things like this.

Behavior like this is completely unacceptable, it is equivalent to store owner throwing you out because he thought your hair are red.

People whose accounts are being suspended without warning, without explanation, and without a way to appeal, like in this case, should be able to sue the company for a large sum of money, so that company knows that breaching contracts is not a good thing.

The job of government is to help people establish contracts between themselves, and to help keeping the explicit and implicit terms of the contract. The random suspension of account is likely a more serious breach of contract than selling private data to third parties.

> Behavior like this is completely unacceptable

This comment gives a lot more context:


IMHO, it seems a lot more reasonable.

No, it's not reasonable.

>The emote spam in question was not "minor", the accounts affected averaged well over 100 messages each, within a short timeframe.

This is a crazy bad take. 100 messages in a few minutes is totally normal for live-streams. Just hold the ctrl-V button and... oops, your G-Mail is locked.

That. Is. Crazy. I worked at Twitch and the idea of disabling someones Amazon account for spamming emotes (even temporarily) is completely insane.

Nuts. Why not just have a cooldown for emote spam? It's ridiculous it's that easy to have your shit banned.

I do understand how difficult it is to deal with spam, and how important it is. I also see that the issue have arised from error/negligence and not from malice. But blocking the whole google account including the email, that can contain very important or urgent information, is completely unacceptable.

A small site like HN can block someone by mistake then apologize and say "Yeah, we definitely fucked up there", but google pushes gmail to make it essential for everything: login to phone, login to chrome, built in password manager in chrome, oauth login to third party sites. So it must handle things much more carefully.

I love how their final point is "there's just too much content for us to manage".

Google, you're being paid by advertisers for that content. You hold people's entire livelyhoods in the palms of your hands. No individual's ability to access their email should depend on the number of emoticons they're putting into a chat channel.

Step up and put a bit more of your billions of profit every year into making sure you're not destroying lives by empowering poor automation.

Exactly, and even if they do not want to spend billions to manage all the content, they should spend at least a few millions and implement per site blocking + warnings, so that a small error in youtube spam filter results in a small issue of youtube account not being able to post comments.

This is just a common sense, because with their current approach of nuking everything and then sending explanations to the blocked email account they are eventually going to upset either enough people or someone wrong, and that will cost them way more.

I was bored and did the math, and found that Google could conceivably hire 10 thousand people and their managers for above the median US household salary, and not break a billion dollars.

A billion dollars has a lot of employment power. And Google makes around $8-9 billion in profits.

Doesn't seem reasonable at all?! Youtube's response is, at the most charitable, being dishonest in this post. They're directly contradicting the video evidence and screenshots. They blatantly directly LIE that the accounts have been reinstated when many have not. It's about as unreasonable as it gets, a corporate communication non-answer would have been more reasonable than straight up lying.

Also, if you're running a service that's grown too large for you to properly manage (which clearly is happening, even to Google), then if you fuck it up, that's still on you! There's no fundamental right to build a really large thing, if you can't manage it properly. So they thought they could get around this limitation with AI and robots, but that doesn't work well at all. That doesn't mean you can go "oh well we're trying, but there's just too many videos being posted!" yeah well DUH that's because you keep on growing while not having the resources to keep up with the videos being posted.

It's like I built a skyscraper but people keep getting hurt in it cause bits are falling off and the floors aren't strong enough. I can't just go: yeah I built this huge skyscraper and I really TRIED reinforcing the floors with duct tape but it just doesn't work well enough. There's really nothing I can do, there's just too many floors, we're doing the best we can and it's unavoidable that some people will (literally) fall through he cracks.

That's not reasonable, I probably just should give up on the skyscraper or sell it to someone who can take responsibility for it.

> without explanation

Common misconception. Let’s assume something as bad as a criminal in prison. Well, they still need their emails. Perhaps not in prison but at least for their recovery. Or to access their bank account. Their retirement funds. Their social security. It’s a misconception that there could be a good reason to entirely revoke any account.

Youtube is different. It’s a broadcasting tool.

There can be some special cases when suspending account is justified, e.g. criminal not in prison using the account for criminal activity, someone creating new account and sending spam from it or collecting phishing data. So we do not want to completely prevent companies from banning users who abuse the service and cause financial damage to the company, we want companies to not mistreat the normal users who happened to trigger a bug in some algorithm.

But i was not saying that explanation alone would make suspension ok, i was saying that the lack of explanation is an aggravating circumstance in this case.

Correct, they shouldn’t be allowed to keep using digital means to perform criminal activity. Perhaps a read-only email or receive-only would be an idea. I just wanted to make people think twice about the idea that suspending an account based on ToS is necessarily as legitimate as currently accepted.

Just absurd, the damage of having my entire google account being banned is seriously non-trivial and would easily cost thousands of Dollars worth of time and energy to deal with. I remember having the fear of what's happening here when Google originally purchased YouTube but after years of non-issue I forgot about the possibility. For me, this is the last straw with Google, the risk of being banned for such an arbitrary reason is simply too high for me and I'm going to start moving documents and accounts from their services.

It baffles me, they hire some of the worlds more intelligent people but the products/systems they produce are constantly idiotic and daft.

There are too many people here advocating to host your own email in your own domain. Please be warned that it has its own risks:

By hosting your main email on a custom domain(not a provider) you open a new attack vector for identity theft. There was an article on HN just a couple weeks ago about someone getting hacked by exactly this attack form. IIRC it was godaddy having stupid verification process.

Someone can hijack your godaddy/namecheap/gandi account and point the MX DNS records of the domain to their own server and receive all your "Forgot your password? here is the link to reset!" emails

This a very bad advice unless you actually know what are the risks.

Maybe suggesting a paid email provider would be better, does anyone know of any reliable email providers that you pay for what you are getting and they are not selling your account data or block you for some reason?

Also, you randomly won't be able to send mails, as google will randomly mark them as spam (for no reason, with no explanation and with no chance to appeal)

So you'll never know if any mail you send to a gmail user (which is pretty much everybody) will ever actually reach an inbox



This. I recently found that even though several emails from my custom domain accounts had gotten through to gmail users - some of them did not get to the main box instead ended up in promotions.. and some of them did not get delivered at all.

These cases were to the same person, from the same computer and account - some would go through and some would not, some could be found in spam box, others just vanished -

then I started calculating the time I spent writing some emails that I never got a reply to - like hours of writing - and I though I was just ghosted for a competitor - turns out it's likely they never saw the work I sent.

I have witnessed this with 3 different receivers the past year.

The last back and forth I just resorted to using my personal gmail account to communicate with new business associate as the receiving was so erratic. At first they wondered why google was putting stuff in spam folder that did not belong there - then it just turned to non-stop frustration and delayed thumbs up on work orders.

the algo should know if you sent one email and they opened and read and replied - and you sent another, it should not matter if you put in the second email, 'your desire to do Z with your web site is understood, and can be done, but you may run into a bad SEO issue that could affect where your site displays in search results, so instead we reccomend not doing X and instead doing Y.."

something like that - so many times is seems that the anti-spam seo hating google teams trump and destroy all the things, regardless of false positives - and no transparency which hurts people - but saves the secret sauce of the anti-spam team I guess.

This is a tough place to be in when so many use gmail.

Constant news about Google banning accounts for various mistakes, small or large, reminds me of a certain country's social credit system. When will they start banning people that have any kind of contact with accounts in bad standing? Will they start denying purchases on Google Pay depending on how well behaved the account is?

Companies are authoritarian dictatorships

Which is ok if they are small, have many concurrents, and their field of activity is old enough to have reasonable laws. Unfortunately for google none of this is true for them. And if they don't want more people demanding to break them up, or proposing unreasonable laws for internet, they should be smart enough to not do stupid things like this.

This is a good model. Companies have two constraints keeping them in line: if they cease making money, they die (in general, at least); and the law can tweak their behavior toward benevolence.

Which is why its baffling to me that lots of people who are deadly afraid of government tyranny are completely ok with privately owned tyranny (even monopolistic kinds)

Government tyranny is a far greater danger because government can use force in the form of army and police, and when established it effectively owns everything else in the country.

The privately owned tyranny has much less power, and when it becomes bad, government can intervene.

So while no tyranny is ok, government deserves much more attention, because it is the more powerful tool, that needs to be protected much more vigilantly.

Some of these private multinational corporations are easily more powerful than some smaller governments.

Let's just agree that both can grow bad. It's just that clearly, governments+people can grow a LOT bigger (in absolute number of people) than a corporation, before it generally goes bad.

None of the largest corporations are really trustworthy, and neither are the largest countries stellar examples of good governing. The difference, however, is an order of magnitude more people being "managed". Apparently Google (Alphabet) employs about 100,000 people, most of them subjects not rulers. That's just a very small city or large town.

If we compare this to a small-medium sized country with a reasonably well-working economy, the difference becomes laughable: The corporation is a much, much, smaller group of people, with a disproportionally unwieldy huge amount of power, completely unable to control themselves and behave in an ethical manner. It's pathetic. YES they absolutely need a government to regulate them just like a baby needs someone to change their diapers.

> It's just that clearly, governments+people can grow a LOT bigger (in absolute number of people) than a corporation, before it generally goes bad.

I don't think it's so easy to separate the two. A corporation is not a real thing, it's a legal fiction created by a government. If a corporation acts badly, it's really the government that allowed it to act badly.

The complications occur when a corporation obtains enough power to be able to influence the government, and thus the ability to influence the rules by which it abides, like a recursive program which is able to alter the mechanics which created it.

At this point, the difference between government and corporations start to blur. It's entirely possible for a corporation to be initially created by a democracy at the will of the people, but over time, as its size and power grows, is able hold the government captive, and control its own rules.

Sadly, this is not just a hypothetical thought experiment.

The people who control the people with the guns are the government. In the west, those people are somewhat separate from e.g. private companies, but that's just how we choose to do things, not some intrinsic feature of governments.

What blows me away is that the supposedly manual reviews believe that these Google-wide permanent bans are appropriate for “spamming emotes” in a stream.

Google employees, you hold too much power to be so arbitrary with your bans. You hold people’s digital lives (and, with Android, a portion of their physical lives) in your hands. Please be a bit more responsible with that power.

And he'll make a fuss, a video, and it'll get fixed. They'll all get their stuff back. It'll add to the\ sour taste in his mouth but it'll get fixed.

While smaller creators in similar situations are completely out of luck. Throw a support ticket into the ether. And just have to hope more and pray a bigger Youtuber picks up their plea and their video gets popular so that it gets fixed.

This boils down to trust. We need to trust that our email provider won’t turn off. We need to trust that our ISP won’t shut off our internet. We need to trust that our cell provider won’t shut off our service. We need to trust that our power company won’t shut off our power.

There is simply far too much on the line here to not have protections. The use contract of email providers needs to regulated.

I've been working on a theory for a few years now that advances in communications technologies fall somewhere on the spectrum of either tending to, or absolutely, reducing trust between individuals and social groups.

Note that trust != reliability. Trust is an extension of faith beyond available information. Reliability is a statistical measure. You can achieve higher reliability with low trust through surveillance, enforcement, punishment, and/or a limited-options environment. But that's not engendering trust (and may actually be a reasonably good summary of why tech tends to break down trust -- I'm still thinking this through).

Keep in mind that trust is a two- or multi-way street. In our mediated relationships, there's trust between user and provider, provider and user, and third parties. Any or all may be or contain untrustworthy elements. And the channels of information are limited making decisions on who or what to trust challenging.

It's complicated. Though I think Google can and should do better.

I wish platforms like Facebook and YouTube would stop pretending they can build “communities” with one billion people.

Yes communities is the wrong word. They are just YouTube viewers, the people watching are not united in a common purpose and may be watching for many different reasons. Youtubers are just mass market celebrity like it's always been. Some of them are barely 'creators'.


"Google may also stop providing Services to you, or add or create new limits to our Services at any time."

Wow I'm sure that they are compliant with the EULA is a relief to everyone involved.

Google’s approach to coping with bad actors, i.e. immediately cutting off people from their entire Google presence for anything that trips their automated systems on any individual platform, is incredibly abusive. People can have years and years worth of data locked away permanently simply because they post a line of emojis on a YouTube video? Simply astonishing.

Why would anyone trust Google with any important part of their lives based on this kind of behaviour?

Things like this are a good argument for breaking up the tech giants. It's an example of how monopolies hurt consumers.

Wonder if the ban related to this rule

Repetitive comments: Leaving large amounts of identical, untargeted or repetitive comments.

Looks like they are trying to use AI for automation and it's screwed up. Again. And it will and must happen again and again to get better.

Actually they call this stressful work and actively explicitly are working towards not getting workers to do this work for their own benefit if you believe HR and the mental health supporters of this policy.

I forget the actual term it's something like "Mental Health Ops'.

Why should Google make their contactors suffer digging through video chat when they can get an AI to do it for them? This is literally their policy, it's not satire.

Yes, but just block further commenting, lol. If your algorithm has bad false positive rate, don't hook it to "delete this person from Google" action, but something less intursive.

> Looks like they are trying to use AI for automation and it's screwed up. Again. And it will and must happen again and again to get better.

Excuse me? Why can't they get better data?

It seems here they used email spam data to mark chat from stream. Or something similar along the line.

If you ever watch Twitch livestream people spam, it's encourage unless it's a oneman spam.

Twitch system is much better. Even if it's subjective I'd take that over what Youtube currently have right now.


Also they could validate their model better before rolling it out in production and ruining people lives.

"The accounts have already been reinstated. We handled that last night. 2. The whole-account "ban" was a common anti-spam measure we use. The account is disabled until the user verifies a phone number by getting a code in an SMS."

Is g00gle harvesting mobile numbers?

1. Yes obviously

2. They are lying through their teeth about the accounts being reinstated (many have not been), as well about the SMS thing, given that the video shows actual screenshots of people being denied having their accounts reinstated, so obviously that SMS thing doesn't work at all as they claim it should.

Google seems to be constantly playing with any of their algorithms... I totally understand that it’s their business, but especially YouTube gets worse and worse.

oh wow! there must be more backstory to this. finding it hard to believe that whole google accounts got suspended for spamming a youtube video chat!

This is why I avoid as many google properties as possible. Who even knows what actions might get my Gmail account disabled. And when they do, there’s no chance of their support system fixing it. The ban bot decided a rule was violated, and when reviewing its decision the ban bot unsurprisingly decides the same thing on appeal.

So no, I’m not going to start shopping with Google Wallet no matter how many coupon offers they send me.

And of course the not-stated-but-there "You can never use an Android phone again, switch to iPhone or a feature phone."

Sure you could create a new Google account and use that with your phone, but Google's going to know it's the same person (same phone number, places where you have to enter your name, same credit card info if you're purchasing any apps on the Play store, same physical location and movement patterns, etc.) and can simply ban you again. You'd like to have Chrome syncing things between a desktop and laptop? Nope nope nope.

I would not rely on a third party (especially a platform with so inconsistent behavior and unclear rules) when talking about an irreplaceable asset.

Yeah, switching to a better email provider is on my to do list, but it’s a big project that I’ll get around to “someday”. Easier to fly under Google’s radar.

I'm sure some of those commenters in the video stream believed the same thing.

They made the critical mistake of touching a second Google product

Apparently emails were sent by Google to support, to their GMail accounts, which they are banned from...

Sometimes there is not much back story. Google has banned entire accounts for far less; not using Google wallet, apps similar to each other, using donation links. You lose everything, drive, Gmail, domain accounts, AdWords, chrome login.. wait until they make Google search only to registered users

Google's management of YouTube has been terrible the past several years. I don't find it hard to believe there isn't more to the story.

Would it be a reasonable and temporary solution to ban accounts at a service level? So spamming on YouTube bans access to YouTube, etc.

How does this even get coded? Bug in some rules engine for detecting spam?

Much of Google's behaviour seems to be AI/ML based.

Which effectively means Google's own engineers don't understand what is happening or why.

You'd think that Google would get better at dealing with this situation after, oh, say, six years of having the problems of forced multi-service integration (and abuse countermeasures) pointed out to them:

https://news.ycombinator.com/item?id=6746731 [1]

But no.

Just from my own experience, I've been locked out of my Google account[2], my real Google account[3], and of course, when the company, as it on very rare an infrequent occasions does, cancels services, found myself (and a few million others) scrambling to salvage content and communities.[4]

Others have had slightly worse authentication/access situations.[5]

To be fair: the problem is a complicated one, and Google sees a lot of abuse. Or, as in Mariplier's community's experience: activity which looks a lot like spam, at a multi-billion-user scale, which can make assessment difficult.

(Though as has been pointed out: if scale of operations makes reasonable levels of service provisioning prohibitive, perhaps you shouldn't be operating at that scale?)

Google do manage to cull a lot of spam, abuse, and other crud. In my some-time career of trying to get hard numbers regarding aspects of Google+ membership and activity, I found:

- G+ communities were being created and deleted at a prodigious clip of thousands per day. Even in the final months of the site, several hundred thousand new communities were added.

- A strong predictor for whether or not a G+ profile would be deleted was if it had ever posted public content. By a factor of about 10 over profiles with no visible content. Presumably, many of these were spam or other abusive profiles.

- The very most active profiles across a small sampling of recent Communities posts were spammers, and by the time I'd gone back to verify the top few, most or all their content was removed.

When you're operating at large scale, it's difficult to make discerning, accurate, fair, or consistent decisions. That's understandable.

But you should be aware of this and design systems to recover gracefully from errors. Which includes not discomfitting, annoying, or distressing users excessively. Someting Google have repeatedly failed at.



1. The day that item ran, the top three items on HN were either directly or indirectly about my experience or frustrations with the then-new Google+ - YouTube accounts merger. https://i.imgur.com/YgEjUuI_d.jpg?maxwidth=640&shape=thumb&f...

2. https://old.reddit.com/r/dredmorbius/comments/2w618r/how_to_...

3. https://old.reddit.com/r/dredmorbius/comments/3mo7l6/that_go...

4. https://social.antefriguserat.de/index.php/Goals

5. https://www.nytimes.com/2016/12/13/us/politics/russia-hack-e...

I wish we could ban them too.


Some black box algorithm deciding you're spamming getting your GMail, Google Drive, Photos, etc. banned for an unknowable period of time, combined with Google's amazing customer service, is a different story than some forum moderator suspending you forspamming.

A comment like this makes it very clear you’ve never engaged with a live streaming community - and you also didn’t even view the source video. The “spam” in this case was specifically directed and requested by the live streamer to help him with crowd sourced decision making.

having fun spamming stuff in livestream chats is very much a part of livestream chat culture.

at the very least, getting your entire Google account banned over it is completely insane.

You mean YouTube comments are now moderated? That’s very welcome and long overdue.

This is about livestream comments.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact