Chess.com mutes sexual abuse victims for talking about their abuse

I've had a bit of a conversation with @Martin_Stahl about this but it is going nowhere so I'm taking it here, in public, where chess.com 's reputation is at stake.

___________________

Here's the story:

So I got in a private message conversation with someone about sexual abuse in religous intitutions and wider society, and we shared some personal stories of people close to us who are affected.

Partway through that conversation, she got mysteriously muted. This only lasted for a short period, but when she got unmuted apparently the reason given was her using a legal term for prolific sexual abusers.

Apparently chess.com has a bot which goes through private messages looking for "slurs" and automatically mutes anyone who have used those words. And the legal term she was using was on that list.

Understandably she was reluctant to talk about that topic afterwards.

And the fact is, if this term is on the chess.com "blacklist" it's also going to trigger when direct abuse victims talk about their experiences.

chess.com is silencing abuse victims.

This should not stand.

Any such words which might be used by abuse victims should be removed from the chess.com blacklist as soon as possible.

Why are/were they talking about it on chess.com, though?

To clarify, when posting content, the process of posting gets checked against the content filters. If terms in the filter are found the system gives a red banner warning. If the content is posted again, it can result in being automatically muted, but normally it's after multiple warnings in a relatively short period.

The system is not going through private message, per se, but checking all content before posting.

A first mute from the auto-mod is 24 hours.

As mentioned in the messages, it's not acceptable for one member to message another member and use those terms perjoritively either and if allowed is likely to cause more harm than prohibiting it. Also as mentioned, the filter isn't something that can determine context so erring on the side of disallowed content is going to be best for the majority of cases

Finally, as I also mentioned, if the member is on your friend list then the filters are not applied as far as I'm aware

I wasn't timing how long it took, clearly that was inaccurate so I corrected. I don't think that affects the substance of my complaint though.

There are clubs which talk about a wide range of topics.

This was a continuation of a conversation we were having in the club, although going on a bit of a tangent.

But people talk about all kinds of different things in private messages anyway.

And I can also imagine it might be relevant if someone sees someone else on chess.com they have some real-life history with.

Wait, chess.com filters DMs? That's dumb. In a public forum I could understand, but DMs?

Abuse happens in DMs

This is in private messages, not in public.

Also, could you at least acknowledge that even if this is something there is no need to talk about; muting someone for talking about it when they could be a irl abuse victim is extremely bad form?

Slurs aren't really a big deal anyway, and in this case you'd only be removing one possible slur from the blacklist.

Potentially traumatising abuse victims is much worse.

I think given the balance needed and explained by Martin I side 100% with chess.com. Abuse of any kind needs challenging, however the filters in place are I am sure stopping a lot more abuse and chances of abuse than allowing. If you and your new friend have bonded perhaps a more private and intimate place would be good? I am pleased you found a kindred spirit on here and I wish you the best of luck.

Direct messages are not being filtered between friends on chess.com, are they?

I'm not talking about removing the content filters altogether (although I do think this is a sign they are a bit over-enthusasitic), I'm just taking about removing one or two words from their blacklist.

According to Martin it doesn't apply for messages between friends, we hadn't friended each other at the time.

I don't think chess.com makes clear to the public when these do and don't apply

Chess.com is absolutely right in using filters. Especially since I once fell for this bait myself. I admitted the mistake and we peacefully agreed on everything with the support. 
but I didn't run and shout about it to the whole cosmos...
Regarding the post - we are stepping on the same rake.
A couple of months ago, the same post and accounts were closed for spam.
In general, Chess.com is for chess, and if someone commits an illegal act - there are law enforcement agencies to resolve this issue.
Martin, please delete this post!

I heard a few years back from @erik himself that if you friend someone, you can say whatever you want in dm's and chat. Is that still relevant today? Just wondering.

That's my understanding

People can also self-censor (such as by replacing one or more characters) to say the same thing without triggering the filter.

Which means that the site's censorship tool will let intentionally abusive people easily circumvent the filter because they are already conscious of "doing something wrong" and thus remember to circumvent it, while people innocently using the wrong word unawares (including: while talking about abuse) will get penalized.

It would be better to simply not have such a filter.

Attempting to bypass the filter may work, the filters can catch a lot of attempts, but when reported the member has no excuse for intent and could get a more severe result.

The filters warn members that something in their post in not acceptable. So abusers that don't get the message will be muted or if they try and get creative will again likely see a harsher penalty when reported.

Just allowing everything and anything isn't a viable solution.

I understand why you see allowing everything as unviable.

I am a little skeptical that automated mutes are necessary: moderators are able to govern the public forums for things other than slurs (such as accusing people of cheating), so why do they require filters to govern PMs and such for slurs specifically?

But let's assume for the sake of the argument, that automated mutes are necessary. Fine. You can keep them.

All I'm asking is to remove the one or two words from the blacklist that might get victims of abuse muted for talking about it.

It would still mean e.g. smokers might be unfairly muted for using a slang word for cigarette and other people might be unfairly muted for using "gang speech". I don't like it but sure. But the risk of traumatising abuse victims is much worse than creating an inconvenience for smokers. You're not talking about the same magnitude of thing here.

I just don't get the refusal to make any changes to the blacklist at all. If you're committed to a blacklist, it's still logical to try and reduce the harm it could cause.

I certainly can sympathize but allowing people to accuse someone of being*one of the filtered terms* or that you wish they get "another filtered* word seem much more problematic to me 🤔

I think this is an American site, and that can sometimes come with a bit of a culture clash for the very many non-Americans here (for example, Europeans and Australians swear very casually in ways that may easily shock a lot of Americans, especially Americans over a certain age).

To the Americans, we Europeans may seem "obscene". To us Europeans, the Americans may seem childish.

With regard to "just allowing anything and everything is not a viable solution", I think that on an automation level, it could be. Because the automated censor already isn't stopping people from being abusive anyway; it just inconveniences people.

If it were left to manual reports only, then people don't get mutes for using the wrong word, and actual abuse can still be moderated approximately as well as it was already.

The "automated filter doesn't check conversations between friends" thing is a little weak, because:

1) when did this happen, and are you sure about it? Because I've for sure been warned about such before now, between friends

2) I run LGBT Club on this site (and have done for many years), and in that capacity I often get messaged by (mostly marginalized and vulnerable) people seeking social and/or emotional support. Which can often involve talking about things that use some words related to sex. These people often don't add me as a friend first. Not only can I thus get warned/muted if I use a "bad" word (and I must guess at what words are and are not allowed), but also, they can too. Which means that not only is it a problem of me, a very seasoned Chess.com member, having to avoid the problem, but also new members who are clueless. And it has happened so very many times that someone comes to me seeking support, and gets muted in their first day on the site. And sure, they (mostly) come back the next day, but it's really a very bad welcome.

3) The problem can also occur in locations such as clubs, club chats, and the like, where who is added to whom as a friend makes no difference.

With regard to my personal motivation about this, I'll be clear:

1) I personally can work within a framework of not using certain words (although it'd be easier if Chess.com provided a list of what they are, instead of leaving us to figure it out by trial and error). I'm even a professional writer, and I have to abide by various things while writing for my mostly American audience at work. Here on Chess.com, I can still forget sometimes because I'm not so focused on my word-choices as I am at work, and unlike at work, where I can edit things 20 times before they go out, here, once I press "enter", if there's a forbidden word there, I can get a warning or worse.

2) However, it's a pain (see, that was not the first word that came to my mind, but knowing the censorship being as it is, I avoided the word that rhymes with itch) for members who don't know yet about the censorship, and/or are not thinking about "I wonder who is screening my private messages when I talk about this emotionally-charged topic". Same also for people who are not even writing in English, but use a machine translation to write it in English, and something that was culturally very normal in another language, uses a forbidden word in English. Same also goes for people who are not using translation at all, and are just writing another language and using a word that happens to be written the same as a "forbidden" word in English (I've fallen foul of this before, myself). This is an example of a case where the warnings are especially unhelpful, as it just says "your post may not meet our community guidelines" and it leaves the writer to try to work out what they said wrong, in the comment that is no longer even in front of them for examination, because the censor ate it.

3) I've been a diamond member on here, give or take a few brief spurts of other membership tiers, for almost as long as Chess.com has been a website. What keeps me coming back here every day, when Chess.com's main competitor is free, is the community aspect, and especially how it enables me to provide support to those who need it, especially in an era when more local community is hard to safely find for a lot of people under the LGBT banner.

So, yeah... It's obviously not the biggest or most critical thing that needs fixing, but if we're discussing it, then at the very least let's say, there's considerable room for improvement in order to make this (as presumably the intention is!) a pleasant environment for everyone, not just for middle class straight white Americans over a certain age.