Skip to content

The "Go-Next" Dilemma: Why AI alone & Team-Votes suck – and a simple Hybrid-Idea (CS:GO Overwatch st

Mrbecke Mrbecke
in Feedback and Suggestions

Hey everyone,

We all know the pain. Someone wants to "Go Next" (give up immediately), the automated system acts up, bans innocent players who were just getting tunnel-visioned, and everyone is mad.

Why? Because:

  • Pure AI is dumb: A strict algorithm has zero context. It can't tell if someone is intentionally trolling or just getting brutally farmed by a sweaty killer.
  • Pure Team-Votes are chaotic: If your team is angry right during the match, emotions run high and people abuse the system out of pure salt.

So, how about we stop relying on extremes and use a Hybrid-System (kind of like CS:GO's old Overwatch)?

Here is the dummy-proof breakdown:

  1. The AI Flags It (The Detective): An automated filter spots suspicious "giving up" behavior after a match ends. It doesn't ban anyone by itself—it just marks the clip. No false bans on the spot!
  2. Cooling Off (The Time Delay): The flagged short replay is reviewed after the match. By then, everyone has cooled down emotionally.
  3. The Neutral Jury (The Overwatch): Instead of salty teammates, neutral players (or a rotating pool of players) review the short clip out-of-match to see if it was actual intentional trolling or just bad luck.
  4. The Veto Rule: If there's any doubt, or if someone votes "Not Guilty", the player walks free. No unfair penalties.

TL;DR: AI finds the needle in the haystack, neutral human review checks it with a cool head after the match. No more broken auto-bans, no more emotional team reports.

What do you think? Let’s hear your thoughts (even if you think it's garbage)!

Comments

  • Arcturus Arcturus

    How would you get random players to review these clips, if that's even possible? At least as of now replays are impossible because the server doesn't actually record all of the information for a replay

  • unknown unknown
    This content has been removed.
  • DodoSwarm DodoSwarm

    @Mrbecke said:

    Hey everyone,

    We all know the pain. Someone wants to "Go Next" (give up immediately), the automated system acts up, bans innocent players who were just getting tunnel-visioned, and everyone is mad.

    Why? Because:

    • Pure AI is dumb: A strict algorithm has zero context. It can't tell if someone is intentionally trolling or just getting brutally farmed by a sweaty killer.
    • Pure Team-Votes are chaotic: If your team is angry right during the match, emotions run high and people abuse the system out of pure salt.

    So, how about we stop relying on extremes and use a Hybrid-System (kind of like CS:GO's old Overwatch)?

    Here is the dummy-proof breakdown:

    1. The AI Flags It (The Detective): An automated filter spots suspicious "giving up" behavior after a match ends. It doesn't ban anyone by itself—it just marks the clip. No false bans on the spot!
    2. Cooling Off (The Time Delay): The flagged short replay is reviewed after the match. By then, everyone has cooled down emotionally.
    3. The Neutral Jury (The Overwatch): Instead of salty teammates, neutral players (or a rotating pool of players) review the short clip out-of-match to see if it was actual intentional trolling or just bad luck.
    4. The Veto Rule: If there's any doubt, or if someone votes "Not Guilty", the player walks free. No unfair penalties.

    TL;DR: AI finds the needle in the haystack, neutral human review checks it with a cool head after the match. No more broken auto-bans, no more emotional team reports.

    What do you think? Let’s hear your thoughts (even if you think it's garbage)!

    I kind of prefer the Behavior Score implemented into Dota that effectively lets every player in the match have a say on whether or not someone was intentionally doing something toxic or abusing a mechanic from the outset and after a moderator views it, the system lowers their overall behavior score (which is between 0 and 12,000 and I have no idea why its that number specifically but it is) and informs both the person being reported and the people who reported them that an action was taken and by extension that affects the type of games they are going to get into. if you want to consistently grief or abuse things then you can find yourself in a lobby full of people who do the same, and only over time does the score go back up.

    (This provides the reporting players with positive feedback that their effort actually meant something rather than leaving them in the dark and also properly informs whoever was reported what they were doing wrong and then they know not to do it if they want to be in respectful lobbies)

    But I think part of the DBD dilemma right now is that people are at odds on what is even reportable or considered griefing.

Categories