Genie Bot & WTF

[chrstgtr]

GenieBot is muting people for saying wtf, which is a common abbreviation for will to fight. 

[Stan`]

@rossenburg

[rossenburg]

6 hours ago, chrstgtr said:

GenieBot is muting people for saying wtf, which is a common abbreviation for will to fight. 

@chrstgtr the bot doesn't have permission to mute users yet. It does send a report to moderators and lobby helpers whenever it detects profane or  pejorative in the chat. It also alert players not to use that word. 

Incase you got muted by any means, it is not related to the bot but instead a moderator or lobby helper took that action.

Could be several reasons:

• wtf itself shouldn't be used in the lobby as its obviously "what the f*ck" not "will to fight" . If you are getting reported for typing wtf when it you meant will to fight then its advisable to type the full word
• Spam: the bot itself has cooldown features which allows it to decide if it has already responded to what a user said within a certain duration and stops responding to the same message until the cooldown is over

Upon checking, the mute action was taken by a moderator because wtf is considered "what the f*ck" and has nothing to do with the bot itself. The bot will of course alert moderators whenever there are profanes in the lobby. 

Thanks for the report, don't forget to report any unusual activities in the lobby as the bot is in alpha stage, we will continue to work on it

Edited December 20, 2022 by rossenburg

[wowgetoffyourcellphone]

Are we really banning "wtf"? 

[Norse_Harold]

1 hour ago, wowgetoffyourcellphone said:

Are we really banning "wtf"? 

If a person uses "wtf" to clearly mean will to fight then it's allowed. If a person uses it to mean "what the f*ck" then it's being added as an infraction. Enough infractions and the person gets muted for an increasing amount of time, starting with 5 minutes.

In the circumstance with [GenieBot] last night, Akazid, nz and Rauls were abusing it by triggering its responses over and over. Until the delays between allowed responses can be tuned better with [GenieBot], players need to be conscientious and avoid producing excessive output from the bot.

[wowgetoffyourcellphone]

Kafkaesque 

[chrstgtr]

5 hours ago, rossenburg said:

Upon checking, the mute action was taken by a moderator because wtf is considered "what the f*ck" and has nothing to do with the bot itself. The bot will of course alert moderators whenever there are profanes in the lobby. 

 

Hard to understand how "wtf == will to fight" gets interpreted as "what the f*ck"

Either way, it's pretty clear that this will lead to a lot of false hits

[rossenburg]

9 minutes ago, chrstgtr said:

Hard to understand how "wtf == will to fight" gets interpreted as "what the f*ck"

Either way, it's pretty clear that this will lead to a lot of false hits

i could also say "sh*t" == Ship High in Transit so im free to use the word Sh*t in the lobby then. I personally wouldn't count wtf as bad word as far as its not spelt out. If its agreed to be used in the lobby, why not. We follow protocols.

Edited December 20, 2022 by rossenburg

[real_tabasco_sauce]

Another thing is that it seems to consider normal conversations spam. Whatever character threshold that is used for considering something spam may need to be increased. I saw a warning go up for a sentence of about 30 to 40 characters.

[chrstgtr]

58 minutes ago, rossenburg said:

i could also say "sh*t" == Ship High in Transit so im free to use the word Sh*t in the lobby then. I personally wouldn't count wtf as bad word as far as its not spelt out. If its agreed to be used in the lobby, why not. We follow protocols.

Sure.

How is that organic, though?

I don't think it is.

That would be an excuse for other behavior. 

Edited December 20, 2022 by chrstgtr

[hyperion]

2 hours ago, wowgetoffyourcellphone said:

Kafkaesque

What is really kafkaesque is the issue as was recently demonstrated in the forums. And if people do something in an attempt to fix this they get criticized or even attacked.

 

1 hour ago, chrstgtr said:

Hard to understand how "wtf == will to fight" gets interpreted as "what the f*ck"

It's very obvious that all those mentions weren't in context were will to fight would be used in abbreviated form normally. Seriously just get used to use w2f in future.

 

58 minutes ago, rossenburg said:

If its agreed to be used in the lobby, why not. We follow protocols.

It doesn't really matter where you set the bar, just handle it consistently. That being said, the list of offensive terms or it's source should be publicly maintained.

[chrstgtr]

5 minutes ago, hyperion said:

It's very obvious that all those mentions weren't in context were will to fight would be used in abbreviated form normally. Seriously just get used to use w2f in future.

 

@Dizaka literally just wrote "wtf == will to fight." I don't know how it is "obvious" that that isn't meant to mean will to fight--it is literally spelled out. 

 

[hyperion]

3 minutes ago, chrstgtr said:

@Dizaka literally just wrote "wtf == will to fight." I don't know how it is "obvious" that that isn't meant to mean will to fight--it is literally spelled out. 

The sole purpose is obviously to trigger the bot, this really feels like arguing with my 10 years old nice.

[Norse_Harold]

The lobby Terms of Use state that one must "not post profanity, pejorative terms or pornographic content," among other requirements. The list of offensive terms is therefore every profanity word in every language and every pejorative in every language. Trying to list them all is not a useful way to spend the time of the lobby moderators, and would make it easier to bypass the automated filters.

Moderators have been ignoring mild pejoratives, profanity, and expressions of violence like "heck", "hell", "darn", "lame", "bummer", "you died", "I killed you (in the game)", "nub", "noob", "nubbed". Basically, think of the kinds of interjections that a 13-year old in a polite household would be allowed to say. Remember that our lobby allows users as young as 13 years old to participate.

Edited December 20, 2022 by Norse_Harold

[rossenburg]

1 hour ago, real_tabasco_sauce said:

Another thing is that it seems to consider normal conversations spam. Whatever character threshold that is used for considering something spam may need to be increased. I saw a warning go up for a sentence of about 30 to 40 characters.

still in alpha development, the bot uses simple regex expression to check repetition of same words in a sentence more than 3 times ( it's not the best way ), and consider it as spam. example typing "aaa aaa aaa" will be flagged as spam. There are more better approach to making the spam filter more accurate.

[chrstgtr]

1 minute ago, hyperion said:

The sole purpose is obviously to trigger the bot, this really feels like arguing with my 10 years old nice.

How do you suppose we learned that it was triggered by "wtf"? And, why would you say the bot's erroneous reaction was correct? 

This is simply an instance where there is an over-broad net being cast, which is why I reported it. 

[Norse_Harold]

1 minute ago, chrstgtr said:

How do you suppose we learned that it was triggered by "wtf"? And, why would you say the bot's erroneous reaction was correct? 

This is simply an instance where there is an over-broad net being cast, which is why I reported it. 

Yes, thanks for reporting the issue on the forum instead of playing with the bot and spamming chat like Akazid, nz and Rauls did. Defc0n (rossenburg) told me that he has already made adjustments to how often [GenieBot] responds to profanity in the lobby.

Edited December 20, 2022 by Norse_Harold

[rossenburg]

23 minutes ago, hyperion said:

It doesn't really matter where you set the bar, just handle it consistently. That being said, the list of offensive terms or it's source should be publicly maintained.

I agreed.

[chrstgtr]

3 minutes ago, Norse_Harold said:

Yes, thanks for reporting the issue on the forum instead of playing with the bot and spamming chat like Akazid, nz and Rauls did. Defc0n (rossenburg) told me that he has already made adjustments to how often [GenieBot] responds to profanity in the lobby.

Thanks—appreciate it

For what it’s worth, I think the bot is good because it explains why things happen. 

[Dakara]

all use insult in french or spanish and it good

[Norse_Harold]

1 minute ago, Dakara said:

all use insult in french or spanish and it good

No, moderators have been noticing insults that they recognize in those languages. Can you provde the moderators with a better list of words to flag, please?

[Dizaka]

Can we just rename will to fight?

[chrstgtr]

16 minutes ago, Dizaka said:

Can we just rename will to fight?

Yeah, let's change to "strong hitting, incredible targeting."

[Philip the Swaggerless]

1 hour ago, chrstgtr said:

Yeah, let's change to "strong hitting, incredible targeting."

I laughed before I even caught the acronym.

[wowgetoffyourcellphone]

On 20/12/2022 at 8:08 PM, chrstgtr said:

Yeah, let's change to "strong hitting, incredible targeting."

Should this change go into the Community mod or on Phab? Please advise.