been thinking about this since my account recently received a perma-ban, and in thinking about it I had never actually had a gauge on how I was acting because there’s no feedback system that lets you know “hey you were reported for this reason” without that there’s no way to correct an issue if you don’t know there was one to start.
OnlyYouCanHearMe2 votesWell, the individual punishment tiers (chat restrictions and the 14 day suspension) are the warnings and the feedback system that lets players know that they have an issue that needs to be corrected. They are accompanied with a Reform Card and chat logs for each punishment showing specific games in which the player displayed actions that earned them a report.
Awf Meta1 votesWell, the individual punishment tiers (chat restrictions and the 14 day suspension) are the warnings and the feedback system that lets players know that they have an issue that needs to be corrected. They are accompanied with a Reform Card and chat logs for each punishment showing specific games in which the player displayed actions that earned them a report.
The chat logs display all of your chat. Even the non punishable chat. It is also possible to get a chat log for a game play offense.
On top of that, "negative" games are held against your account without notice. So, here is an example of a common situation:
Player is mildly toxic. Riot doesn't tell them or punish them. Player continues behaving the same way (why change with no warnings). Some mildly toxic games later, Riot hits the player with a chat restriction and chat logs. The player was breaking the rules for "some" time, while not being warned.
Umbral Regent1 votesI think it's only fair that even the non-punishable parts should be shown in the chat logs. For one, showing only the punishable parts would likely lead to accidentally reinforcing the idea that the IFS runs exclusively off of keywords and phrases. And, that'd only cause more confusion.
Beyond that, though, having the non-punishable parts shown is probably better overall, as in most cases, the punishable parts of the chat should read clearly against the contrast of the non-punishable parts.
As for the negative games being held against the account - you're looking at it the wrong way. It's really more "bad match forgiveness" - everyone has a bad game every now and again, and sometimes, they'll slip up and get aggravated in the chat. These matches aren't held against the account so much as they are being dismissed and noted in case the behavior crops up again.
And, in the case of this:
It is also possible to get a chat log for a game play offense.
Don't quote me on this, as I've never been punished for a gameplay-related offense, but I'm fairly certain that the IFS pop-up tells the punished player that they were punished for a gameplay-related offense. I could of course be wrong, but considering some folks have posted here on the boards mentioning that they "got banned for feeding" and so on - I feel fairly confident in that possibility.
PapaCapone1 votesI have never seen a pop-up in which i was notified i was reported, not one. However i have received a pop-up in which it stated that a person i recently reported had disciplinary actions taken against them. only time i was notified about my punishments was when the pop up said 14 day suspension and perma-ban...
Umbral Regent1 votesDon't quote me on this, as I've never been punished for a gameplay-related offense, but I'm fairly certain that the IFS pop-up tells the punished player that they were punished for a gameplay-related offense.
Bolded and italicized - I pretty specifically mentioned the player getting punished. I don't know (although I do doubt) if a pop-up exists to warn players about their bad behaviors, but I do know that players who get punished do receive a pop-up telling them that they were punished, and my statements here are in assumption that those pop-ups also tell the "why".
PapaCapone1 votes[{quoted}](name=Umbral Regent,realm=NA,application-id=ZGEFLEUQ,discussion-id=q6zfzVbr,comment-id=00000000000000000000,timestamp=2018-02-23T23:30:30.257+0000)
Bolded and italicized - I pretty specifically mentioned the player getting punished. I don't know (although I do doubt) if a pop-up exists to warn players about their bad behaviors, but I do know that players who get punished do receive a pop-up telling them that they were punished, and my statements here are in assumption that those pop-ups also tell the "why".
the player receives a notice that they were punished... that is not what i proposed... outside of the time you actually get punished you have no knowledge of ever being reported, unless you were trying to get reported, so what i had suggested was something to show people where their behavior was taking them by saying "you were reported x amount of times for x reasons"
Umbral Regent1 votesAllow me to also state that the comment of mine you responded to was in response to what FailedAbortion said.
That said, though, people have suggested for personal account status cards before - it seems similar to what you're suggesting, at any rate.
PapaCapone1 votes[{quoted}](name=Umbral Regent,realm=NA,application-id=ZGEFLEUQ,discussion-id=q6zfzVbr,comment-id=0000000000000000000000000000,timestamp=2018-02-23T23:45:54.697+0000)
Allow me to also state that the comment of mine you responded to was in response to what FailedAbortion said.
That said, though, people have suggested for personal account status cards before - it seems similar to what you're suggesting, at any rate.
felt like i had to explain because the very first person to comment didn't get what i meant at all so my apologies.
OnlyYouCanHearMe1 votesWould you consider the ire of player's teammates to be a warning? Such as them becoming increasingly volatile, unable to focus, threatening to afk, all because of the things player says, The comments in both team and all chat "Just mute and report Player." That is about as obvious as it can get in a match that a player is doing something that pisses off the other people in the game.
Awf Meta1 votesWould you consider the ire of player's teammates to be a warning?
Yes and no.
Yes, usually people will show signs of annoyance. A perceptive person would be able to pick up on those signals.
No, if a player is not breaking ANY rules or the accuser lies.
The community tells Riot what they want punished. Riot usually listens. The ultimate decisions are up to Riot alone though. Therefore, Riot should inform the players when they are breaking the rules that Riot enforces.
OnlyYouCanHearMe1 votesIf they are not breaking any rules, or the accuser lies, it won't matter if they are reported. The report is false and it will be thrown out by the IFS.
But as a general rule, if a player is getting in an argument with another teammate that lasts more than a few lines, and degrades into any form of personalized insults, even if their position in the argument is correct, they are breaking the rules of the game. They've moved into verbal abuse, and they are creating a hostile, negative environment where the other players in the match have to suffer for the actions of both parties involved in the argument. And at that time, if someone happens to say "just mute them and report them after" it should be a pretty clear warning to that player that they are doing something that they are going to end up being reviewed for.
The community tells Riot what they want punished. Riot usually listens. The ultimate decisions are up to Riot alone though. Therefore, Riot should inform the players when they are breaking the rules that Riot enforces.
This isn't entirely correct. The IFS was designed and implemented by Riot, but it's a learning system. A type of AI that was created to monitor the actions that the players in each region find inappropriate and punish the players who take those actions; ranging from various verbal filters to the FeederBuster module. With regards to many things, the decision for it to be punishable was made by the players, seen in the overwhelmingly high number of reports submitted. If the community decided today that it wanted to pick a random word and turn it into a bannable offense, it could, simply by having everyone mass report anyone who used it. Within a short period of time, the system would register that the playerbase finds the word offensive, and would begin to punish for it.
Of course, as you say, Riot has the ultimate decision making authority, and can at anytime reprogram parts of the IFS, or override any punishments that were issued. But they designed the IFS under the same principles of player input that the old Tribunal had. It simply doesn't seem quite as apparent, since we don't get to sit down with a bowl of popcorn and read the entire chat log anymore.
Awf Meta1 votesBut as a general rule, if a player is getting in an argument with another teammate that lasts more than a few lines, and degrades into any form of personalized insults, even if their position in the argument is correct, they are breaking the rules of the game.
That is too lenient compared to the current system. In the current system, people aren't allowed to argue AT ALL. While people aren't getting punished for just arguing, the offense stays on record. Possibly, leading to too many unknown minor offenses stacking up.
And at that time, if someone happens to say "just mute them and report them after" it should be a pretty clear warning to that player that they are doing something that they are going to end up being reviewed for.
Notice the accuser doesn't make the rules. The report gets reviewed by Riot. The ire of your team does not influence Riots rules.
Ire is a good warning for account review. Terrible warning for account status.
The IFS was designed and implemented by Riot, but it's a learning system.
I've seen videos on the subject. Learning system is pretty generous in my opinion. IFS is a chat filter. If I were to make my own custom chat filter, it would be "learning" which words to filter. If the IFS were able to automatically and accurately "learn" without player reports, that would be some schweet AI.
OnlyYouCanHearMe1 votesThe IFS is not simply a "chat filter." Whoever told you that, or said so in a video, was incorrect. It is defined as a learning system by the Rioters who designed and coded it, based on the fact that it learns from the reports submitted by players in each distinct region what they find to be unacceptable, and begins to punish accordingly. For example, the phrase "kys" was not always an instantly bannable phrase. It only became so harshly punished in response to the overwhelmingly negative response from players reporting its use. For another example, the word "c - unt" was not a censored or bannable word a few years ago. It was, and still is, considered popular slang in certain parts of Europe, but because of player response, it is now considered a form of sexist hate speech on the NA server. These are things that the IFS has learned in response to submitted player reports.
The IFS is not just some keyword detector that goes through a giant list of all the possible nasty things a person can say to another. It contains many modules, only one of which is a dedicated keyword filter for things like hate speech, death threats, etc. It also encompasses many other functions, including trolling and intentional feeding, neither of which can be detected with a chat filter. Feel free to look through Riot Tantram's posts in this section of the Boards. He wrote the programming for the IFS and discusses how it works on a pretty regular basis.
Awf Meta1 votesIt is defined as a learning system by the Rioters who designed and coded it
Riot and I have different definitions is all.
based on the fact that it learns from the reports submitted by players in each distinct region what they find to be unacceptable, and begins to punish accordingly.
In my opinion, that is not true learning. That is just following orders. The IFS needs to react independent of reports before I would consider it a learning system.
Example for clarity:
Kids taking a test in school. One kid gets to use the textbook while taking the test. That kid is just copying the textbook. I would not say the kid is actually learning.
For example, the phrase "kys" was not always an instantly bannable phrase. It only became so harshly punished in response to the overwhelmingly negative response from players reporting its use.
Do you have a source? I'm pretty sure the "zero tolerance" phrases are hand picked by Riot.
It also encompasses many other functions, including trolling and intentional feeding, neither of which can be detected with a chat filter.
Source? Splitting hairs but I thought IFS was mainly focused on chat and Riot used other systems (leaverbuster) for the game play. Whether the systems are effective at catching trolls and feeders is another topic.
Feel free to look through Riot Tantram's posts in this section of the Boards. He wrote the programming for the IFS and discusses how it works on a pretty regular basis.
Do some digging for Riot Lyte. There are some IFS videos around, it takes more digging than I want to do at the moment. So, I found 1 video. Hopefully, it helps you find others.
https://spark-summit.org/2017/events/combating-abusive-language-in-chat-with-apache-spark/
OnlyYouCanHearMe1 votesRiot Tantram IFS detection modules
Riot Tantram IFS machine learning, not keyword based system
Lyte Intentional Feeding added to IFS
IFS FAQs confirming FeederBuster added to IFS
LeaverBuster is not part of the IFS, but the other gameplay offenses are handled through it. When the IFS first launched, it only dealt with chat toxicity, but within a few months, after players were used to it, inting was added.
Awf Meta1 votesSo, I'll say we are both right. IFS has different sub systems (feeder buster). It is not wrong to consider them all a part of IFS. For this discussion, I am referring to the part of IFS that handles chat behavior.
This is not how the system works. It is not a keyword based system. It also isn't that black and white. It is a machine learning system that learns from the community what is acceptable and what is unacceptable. It doesn't just say, 'they said a word on a list, they get punished.'
At least part of the IFS is keyword based. The hate speech detector. I haven't seen any evidence, I assume the keywords are hand picked by Riot. That is to say, the IFS is not adding words to the hate speech detector automatically.
The real question is, "how does the system learn from the community?".
None of those links you provided gives an explanation. Riot says "machine learning" but never gives an example.
So, here is my guess. Account gets reported. IFS scans. Established "toxicity" gets punished. Possible "toxicity" is archived in the system. If possible "toxicity" reaches a threshold, it is moved to established.
Basically, IFS makes an educated guess. If that guess is popularly reported, it becomes established.
I'd also like to point out, I have not seen anything where the IFS is evaluating non reported accounts. Which means there is no counter weight to this system. There is no comparison between good and bad. Made up example: kids saying "KYS" sarcastically versus kids saying "KYS" insultingly. There is a context difference. IFS would need to understand context to be considered learning.
PapaCapone1 votesi had never received a chat restriction on my account, i got an immediate 2 week suspension and then perma-ban. in the case of the 2 week ban i was being an a-hole i admit i deserved some sort of punishment for my actions but having chat get me a 2 week and a perma ban is ridiculous. i didnt feed intentionally or try and lose the game for my team. after my two week suspension was lifted there was no way for me to tell i was acting against TOS because 1) i didnt know i had been reported and 2) i was receiving honors from teammates for almost 3 months then boom one game off tilt and perma ban... the system i suggested would let you know immediately (or on login) if you were reported and what the reasons were, then if the case was serious enough provide a chat log or what ever evidence and move on to chat restrictions... chat should never get a player banned especially since others can mute you and you can mute them...
OnlyYouCanHearMe1 votesI believe the reasoning Riot has given for why players are not given a running meter of how toxic they are being, or whether they have been reported, is because players would use it to be toxic until right before the point where they would be punished. Then they would back off. And once their account reset, they would resume being toxic again.
It is unfortunate that you were hit with a 14 day suspension and then a permaban. Riot does try to make it very clear to players when they give them the alert about the 14 day suspension that any further violations of the rules will result in a permaban, but many players do not realize that it doesn't have to be a case of extreme toxicity. It's actually easier to go from a 14 day suspension to a permaban than it is to go from a clean account to a 10 game chat restriction.
And I'm not sure if you are aware of this site or not, but you actually can contact Support through Riot Games Privacy and request a detailed copy of your account history. It will include any reports that have been submitted on your account, along with chat logs to accompany them. Just sign in with your normal user ID, and click "Looking for More?"
PapaCapone1 votesI believe the reasoning Riot has given for why players are not given a running meter of how toxic they are being, or whether they have been reported, is because players would use it to be toxic until right before the point where they would be punished. Then they would back off. And once their account reset, they would resume being toxic again.
thats not what i said at all, it wouldnt say "hey you're one report away from a ban" it would literally, as i stated before, say "you were reported for this reason, x amount of times last game."
you cant abuse that...