Lindisfarne wrote: ↑Sat Apr 30, 2022 3:47 pm
Why do you need to mute people for words if you have a filter already built in for ppl who are scared of words.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
Maybe instead of word filter just have automated timeout for ~10 minutes if there's multiple reports against same person? If repeated within certain time period increase the timeout and notify GMs to check it out manually if its legit reports or report abuse by a group.
Word filters never work as people who intend to swear or harrass others will f1nd w4ys t0 g3t 4r0und t-h-e-m, and if you get too aggressie with word det***ion you get fal** ***itives
And dont get me started on the whole different languages part...
Lindisfarne wrote: ↑Sat Apr 30, 2022 3:47 pm
Why do you need to mute people for words if you have a filter already built in for ppl who are scared of words.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
Maybe instead of word filter just have automate timeout for ~10 minutes if there's multiple reports against same person? If repeated within certain time period increase the timeout and notify GMs to check it out manually.
Word filters never work as people who intend to swear or harrass others will f1nd w4ys t0 g3t 4r0und t-h-e-m, and if you get too aggressie with word det***ion you get fal** ***itives
Basically the system could keep on getting entries to a point where variants are completely irrelevant and anyway there will still be GMs to monitor reports. This is a precaution to cover the limited downtime between GMs.
The report-based feature you mentioned could be easily abused, even if there will be retaliation afterwards to the abusers but it's better to just have an automatic system. A good and fair one though, something that is still needed from the current profanity filter as we tweak it.
Lindisfarne wrote: ↑Sat Apr 30, 2022 3:47 pm
Why do you need to mute people for words if you have a filter already built in for ppl who are scared of words.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
I've been trying to get some clarity on this. You said 'The list of flagged words'. Are you implementing a system to stop swearing in the public chats? Is swearing no longer allowed? I asked what the official position was on this, because from what I understood the goal was preventing harassment and hate speech (which doesn't necessarily have anything to do with swearing). I think people are upset because they aren't the type of people who harass other people, but they type the occasional "holy ****" or "what the ****" in region chat. If the new rule is no swearing that's fine, update your TOS and call it a day. Just be clear about it.
Gravord wrote: ↑Sat Apr 30, 2022 12:56 pm
Lowest city that can happen would be 3 stars. Meaning win gives 150 crest and loss 60 crest.
For that to happen same city need to be sieged twice in a row and defenders had to lose first one. If those conditions are not met, defenders won or its different city, it will always be 4 or 5 stars only.
3 star city loss is still almost twice better than 100% sc win rate with perma pops for same amount of time. Again, not counting bags and kills. City win is worth anywhere from 5 to 10 hours of 100% win rate scenarios.
80-100 war crest with 100% sc win rate without taking account of war crest from kill and on top of that a lot more of rr..
Sc win gives 6 crest for 12vs12 sc. Doing 6 of them per hour is already optimistic as some sc can take up to 15 mins and you will not always get instant next pop. 6 x 6 gives 36, not sure how you manage to calculate 80-100 out of it. Also sc have lower crest drop rate from kills than city and rvr by 25%.
That highlights that SC reward crests should be increased. There is even less reason to q city as before the patch.
I only heard 1 person complaining about city crest increase, everyone else was happy with it. I guess two steps forward 1 step back with the time adjustment. 2 hour sweaty city fight with 100 kills both sides for 180 crests vs 2 hours in the lakes getting 300+.
Omegus wrote: ↑Sat Apr 30, 2022 2:21 pm
I'm guessing you missed the angry person spamming racist comments in chat on new accounts all over Easter?
Yep, I was gone for two weeks. So one racist person prompts a system wide change that affects us all? There is already a language filter in place, why resort to more draconian measures? This is a nontroversy.
It was new accounts, every 10-15 minutes, for about a week. GMs (all volunteers doing it in their spare time) having to work around the clock to keep banning the accounts to keep chat clear. This is likely in response to that. It was waaaaaaaay beyond just turning on the swear filter.
Only the public chats are affected by this change, so I think guild/warband/party is fine.
Last edited by Omegus on Sun May 01, 2022 7:33 pm, edited 1 time in total.
Wuu wrote: ↑Sat Apr 30, 2022 6:54 pm
So one racist person prompts a system wide change that affects us all? There is already a language filter in place, why resort to more draconian measures? This is a nontroversy.
The original filter doesn't filter out any racist, sexual or extremely vulgar and inappropriate words and terms. The extra measures were long overdue as we've been interested in such chat filter for years cause the issue with chat spammers is somewhat recurring every year or so, as it's a F2P game and creating a new account and character is pretty easy.
Lindisfarne wrote: ↑Sat Apr 30, 2022 3:47 pm
Why do you need to mute people for words if you have a filter already built in for ppl who are scared of words.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
I've been trying to get some clarity on this. You said 'The list of flagged words'. Are you implementing a system to stop swearing in the public chats? Is swearing no longer allowed? I asked what the official position was on this, because from what I understood the goal was preventing harassment and hate speech (which doesn't necessarily have anything to do with swearing). I think people are upset because they aren't the type of people who harass other people, but they type the occasional "holy ****" or "what the ****" in region chat. If the new rule is no swearing that's fine, update your TOS and call it a day. Just be clear about it.
There was a slight miscommunication and it was simply caused by the default list of words that the system uses as referral. We've looked into over 1,600 words, word by word, to adjust the system's reaction to these words. We will push a fix to adjust the exaggerated moderation of the in-game chat and while some words will trigger an instant mute, others will simply be replaced with stars/symbols just like other games.
We will keep on optimizing the list to maintain a realistic yet safe and friendly environment. The CM team has been closely involved in order to make sure the list and the automatic actions do not poorly interfere with the players experience.
Lindisfarne wrote: ↑Sat Apr 30, 2022 3:47 pm
Why do you need to mute people for words if you have a filter already built in for ppl who are scared of words.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
I've been trying to get some clarity on this. You said 'The list of flagged words'. Are you implementing a system to stop swearing in the public chats? Is swearing no longer allowed? I asked what the official position was on this, because from what I understood the goal was preventing harassment and hate speech (which doesn't necessarily have anything to do with swearing). I think people are upset because they aren't the type of people who harass other people, but they type the occasional "holy ****" or "what the ****" in region chat. If the new rule is no swearing that's fine, update your TOS and call it a day. Just be clear about it.
I'd personally like some clarity on what words are "bad". Like I totally understand blocking hate speech, and I am absolutely fine with that.
My question is what kind of other words? Can I still say 4 letter bad words? Am I going to get chat muted for saying "dammit"?
Like how extensive is this list? If I type "assassin" is it going to pick up the word "ass" and mute me for that? Chat filters that auto mute people like this are always really concerning. Automated systems can be pretty sketchy, and I foresee a lot of appeals flooding the GMs inbox.
The profanity system is being worked on. We would just rather get the system in sooner rather than later (beacuse of recent incidents) and adjust back to what our objective is with it.
The new system definitely needs some tweaking as we mentioned and we will do the necessary changes to ensure a smooth and safe environment. The list of flagged words will be reviewed and optimized.
I've been trying to get some clarity on this. You said 'The list of flagged words'. Are you implementing a system to stop swearing in the public chats? Is swearing no longer allowed? I asked what the official position was on this, because from what I understood the goal was preventing harassment and hate speech (which doesn't necessarily have anything to do with swearing). I think people are upset because they aren't the type of people who harass other people, but they type the occasional "holy ****" or "what the ****" in region chat. If the new rule is no swearing that's fine, update your TOS and call it a day. Just be clear about it.
I'd personally like some clarity on what words are "bad". Like I totally understand blocking hate speech, and I am absolutely fine with that.
My question is what kind of other words? Can I still say 4 letter bad words? Am I going to get chat muted for saying "dammit"?
Like how extensive is this list? If I type "assassin" is it going to pick up the word "ass" and mute me for that? Chat filters that auto mute people like this are always really concerning. Automated systems can be pretty sketchy, and I foresee a lot of appeals flooding the GMs inbox.
At the moment only a handful of people have been muted by the system, so there are no much consequences and appeals yet.
Providing the list of what words are flagged and monitored will most likely not happen, but we're manually reviewing around 1,600 words to make sure they're properly weighted by the system. The worst automated scenario would be a one hour mute for extremely vulgar, inappropriate or racist words while other words like "ass" will be filtered out; Meaning that you won't get muted, but it will simply show the word as ***
While I understand there are some words/phrases you don't ever want getting through, for the... casual swearing?... would you consider moving some of the words to the game's built-in profanity filter instead? Players have the option of disabling that in the game UI settings.
It might help draw the line between allowing players to block general swearing if they want to, while also making life easier for the GMs by having your back-end system try to detect and block abuse, racial slurs, etc.