I'm wondering if Riot is looking for trends in the data with bans they feel are completely justified. In otherwords, this person was found to be in violation of TOS 1, 7, and 12. They also did:
Finding these sorts of things with some multivariate regression analysis might allow for them to better sift through the backlog of complaints, no?
TempName548462 votesthis is how the report system works:
you say: "you fuakin ahtisto boy, just khell ya salf alredy nobody loves you", the system sees: - you being reported in this game results in: nothing
you say: "my little brother was diagnosed with autism yesterday :/", the system sees: AUTISM you being reported in this game results in: chat restricted or possibly perma banned
Starcraft243ver1 votesChamp in a non-traditional lane
Happens a lot recently and it doesnt mean it's suboptimal, just saying :)
Also it's not a violation. It's a violation when the way you are playing is hijacking the fun from your team, while your team disagree on it. (IE : playing double jungle and letting your ADC alone, he's going to be miserable 2v1 botlane).
RookPusher1 votesHappens a lot recently and it doesnt mean it's suboptimal, just saying :)
Also it's not a violation. It's a violation when the way you are playing is hijacking the fun from your team, while your team disagree on it. (IE : playing double jungle and letting your ADC alone, he's going to be miserable 2v1 botlane).
Yeah those were examples. I'm not saying they should look for things, but you'd have to label data for the MVRA to name. Correlation is not causation. Playing a non-traditional champ in top lane might not be trolling. But when trolling, many would play a non-traditional champ in toplane.
Therefore, when that qualifier is met - move on to the next check point. It's just a way of better assessing claims of trolls without being sloppy.
Tegash2 votesHonestly, I know theyre a Small Indie Company, but they could prolly hire somebody to install a neural network and train it up on games.
All the intuition of a human with a fraction of the cost.
RxO Carmada1 votesI'm wondering if Riot is looking for trends in the data with bans they feel are completely justified. In otherwords, this person was found to be in violation of TOS 1, 7, and 12. They also did:
- Champ in a non-traditional lane
- Built items with minimal benefit to said champ
- Tower was taken exceptionally early
Finding these sorts of things with some multivariate regression analysis might allow for them to better sift through the backlog of complaints, no?
The issue is that youre talking about a few thousand people sifting through millions of reports and stuff a day, which is why before they had tribunal where users could punish themselves, and then later changed to have an automated system. There is no plausible way for riot employees to take care of even the most specific instances in which they need to thoroughly review each game. I think the current automated system is terrible, and while I miss the tribunal, if they don't want to use the tribunal than I think they need to rework their punishment system.