Smurfing can make League matches feel unfair, but community detection tools seem to rely partly on people voting on suspicious accounts. How useful is that data, and what are its limits?
Can community tools actually tell who's smurfing in League?
Answers
Community votes can flag accounts worth looking at, but they don't prove an account is a smurf. A player can look unusually strong for all sorts of reasons, and a vote records what someone suspects, not what they know.
There's also the problem of manipulation. One community project reported bot attacks and said it couldn't guarantee that every vote in its database was legitimate, even after cleaning up obvious abuse. I wouldn't treat that database as a confirmed list of smurfs.
Reply to this answer
Detection and enforcement are different problems. A community tool might collect reports, but deciding what action is justified needs more than a public suspicion score. The game operator is in a better position to assess accounts using information a community project may not have.
That still leaves the fairness question. Players want reasonably matched games, but labeling or penalizing the wrong account would create another problem.
Reply to this answer
There's a practical limit too. A volunteer project can end up dealing with bot attacks, abuse and threats just for collecting this stuff. Even if the data is interesting, keeping the tool running and protecting it from interference is a lot to put on one person.
Reply to this answer
Add your answer
Answers are reviewed before publication.
DpadBacktap_6R:
So even a large number of votes doesn't necessarily make an individual accusation reliable?
DpadPivotStep_3V:
Right. A large vote count doesn't tell you whether the votes are independent, legitimate, or based on good evidence.