Report system

SonicAF·10/27/2018, 11:32:40 AM·1 votes·1,292 views

As we came to some sort of understaning in this post, and while I agree that this exact problem is not a major one and the system is never perfect, we can partially solve it and many related ones by motivating people to report responsibly - not "every time somebody makes you uncomfortable" and educate the community in the process.

Nowadays the report system begs to be abused. It's evident from the fact that "report" is one of the most(if not the most) typed words in league, despite the fact that Riot Games clearly expressed that they don't want people to ask for reports, which might be the most specific rule they ever had.

You can report anyone for anything without second thought and the system will try to find a reason, its own one. It's one click away. Some premades do that to each other as a joke without even knowing how it functions, leading to posts about getting punished for intentionally feeding or "zero tolerance" stuff when it wasn't meant to be punished. Hell, you can even report the wrong person by accident.

I think it should be made a simple, but much more thoughtful process, make people actually file the report - not just faceroll through endgame screen. Here is what I propose.


For chat-related offences present people with the full chatlog of the person to be reported. Each message can be flagged to be reviewed for offensive or negative speech.

For gameplay-related offences make people go through the replay and choose time intervals with violations.

People abusing by repeatedly marking the whole game or entire chatlogs should receive a friendly reminder that their reports can't be accepted that way.

People repeatedly flagging innocents should receive an automatic tip that the behaviour they mark is acceptable and/or can't be determined to be punishable with enough certainty.

When IFS learns enough, Riot can start sending the examples of such reports back to the player so he can see what he is doing wrong. There is no need to punish players for that.


I think this would educate people on what is right and what is wrong. Within such moderately time-consuming system we can display tips about everything related to the punishment system, like what is punishable or the fact that you shouldn't ask for reports. I also like the fact that this system's design guarantees that the enemy team can't report someone for the chat they don't see, directly affecting incentive to call for reports.

Just as always, you are welcome to point out what am I missing.

3 Comments

Umbral Regent10/27/2018, 2:30:59 PM2 votes

So, there are a few issues with this idea, first and foremost:

More complicated/involved =/= better. Increasing the commitment players have to take to report someone both decreases the time they're actually playing the game (since they have to file out the report over a longer duration) and it can in and of itself be a turn-off from reporting.

Like, why report for chat offenses when I have to scroll through a chatlog just to pick out the lines of misbehavior that the system already picks up on? Why bother with filing a report for a gameplay offense if I have to re-watch the whole game to set up timestamps for review and punishment? At that point, you may as well give players the executive decision, since it seems you're making them do all the work of discerning the punishable offenses and leaving it up to Riot to simply review for certainty's sake.

The current report system works fine - it's simple, easy-to-use, and it has players file their report and get into their next match within roughly a minute or two, without making them put forth extraneous work that the system already covers.


Another issue is that, in a roundabout way, This idea brings back report weighting to some degree. Granted, it's less of report weighting and more of simply denying the shotgun approach, but here's my question - why deny a report when the IFS can easily sift through the logs and decide whether or not the report is valid?

And, by the way, I'm specifically talking about this right here;

People abusing by repeatedly marking the whole game or entire chatlogs should receive a friendly reminder that their reports can't be accepted that way.

The only point that this function would have would be to commit players to the unnecessary busywork of sifting through logs (which the IFS can already do) and rewatching a bad match they just got out of (which they doubtless don't want to do) just so they can flag specific lines or timeframes.

It denies the shotgun approach, but sacrifices so much for such a small issue that it really does more harm to the idea than good.


This issue is less of an issue with the idea, but more an issue with phrasing/misconceptions;

You can report anyone for anything without second thought and the system will try to find a reason, its own one.

Yes, you can report anyone for anything, but the IFS will not decide its own reason for punishing them - it will review the logs of the reported player(s) and look for behavior that is in breach of the Summoner's Code. It will look for Riot's reason to punish them, not its own. And there's nothing wrong with the system punishing based on Riot's ruleset.


I think this would educate people on what is right and what is wrong. Within such moderately time-consuming system we can display tips about everything related to the punishment system, like what is punishable or the fact that you shouldn't ask for reports.

There are much better ways to spread information than overcomplicating a system that works. Loading Screen tips, for example.

usul120210/27/2018, 4:20:06 PM1 votes

I work part time in a call center. At least 50% of our calls are because people can't read our site/what we email them. It's absurd how even a slightly complex system can completely elude average people who have other shit going on and their brain isn't as focused on what they're doing with us.

While yes, I think some good info could be sent by IDing chat logs, I think that the wealth of bad usage of the system would instantly drown out any gain.

Ultimately, imo, there's no benefit for fomolicating the system. Machine learning works well when you have large sample sizes, which riot certainly does.