The reporting system of Overwatch is just a temporary measure. Soon, toxic players should be automatically filtered out.
Disruptive players who repeatedly insult or oppose their own team are as common in Overwatch as sand on the beach. While players can report these individuals, which eventually leads to punishment, the trust in Overwatch’s reporting system is relatively low and the waiting times are high. However, this is set to change in the future.
In an interview with Kotaku, the Game Director of Overwatch, Jeff Kaplan, talked about the issues with the game’s toxicity. Players often behave inappropriately, and the reporting system is not yet in a perfect state.
Kaplan’s vision is ambitious. According to Kaplan, the team is currently experimenting with learning software. The developers are trying to teach the game what toxic language is and they apparently have a lot of fun doing it. Overwatch should be able to autonomously detect when players behave disruptively and then take automated actions.
The idea behind this is that toxic behavior should be registered before any report comes from other players.
What sounds like a cool idea still faces many obstacles. Because toxic language is not the same as toxic language. While friends may throw around phrases like “You idiot!” or worse at each other without anyone feeling insulted, this can be perceived as a clear insult in a group that does not know each other.
Overwatch should also recognize disruptive gameplay
If this is realized someday, the developers want to go even further:
Overwatch should not only be able to recognize toxic language, but also toxic gameplay, such as a Mei who constantly blocks her own spawn area to hinder her own team.
Initially, when filtering language and recognizing intentionally toxic gameplay, it will only be possible to identify the extreme cases. This is not a problem, as these extreme cases could help further develop the system and continuously improve it.
In the more distant future, Kaplan would prefer not to deal with toxic behavior at all, but rather to implement a “positive report feature”. This would allow players to be “reported” for particularly positive behavior.
Would you find it good if the reporting system were automated and “bad” players were filtered out directly? Or should the final decision always lie with a support team member?
