Skip to content

FAQ

Jump to bottom
Wiki Generator edited this page Mar 18, 2026 · 1 revision

FAQ

General Questions

What is OverWatch-ML?

OverWatch-ML is an advanced anti-cheat plugin for Minecraft servers that uses machine learning to detect x-ray cheats and combat modifications. Unlike traditional anti-cheats that rely on simple heuristics, it analyzes player behavior patterns to identify cheaters.

What servers does it support?

  • Paper 1.21+ (Recommended)
  • Folia (Supported)
  • Spigot (Basic support)

What are the requirements?

  • Java 21+
  • ProtocolLib (Required)
  • 4GB RAM minimum, 8GB recommended

Is it free?

OverWatch-ML is free to use. You can download it from SpigotMC or Modrinth.

Installation

Why isn't the plugin loading?

  1. Make sure ProtocolLib is installed first
  2. Verify you have Java 21+
  3. Check for error messages in server console

How do I install?

  1. Install ProtocolLib
  2. Drop OverWatch-ML.jar in plugins folder
  3. Restart server
  4. Configure as needed

Do I need to configure anything?

The default configuration works for most servers. However, you may want to:

  • Adjust decoy settings
  • Configure Discord webhooks
  • Set punishment levels

Detection

How does x-ray detection work?

The system monitors:

  • Ores mined vs blocks broken (ore ratio)
  • Mining patterns and trajectories
  • Time between ore discoveries
  • Look direction when mining
  • Comparison to ML models

Can it detect other cheats?

Yes, it includes:

  • Combat cheat detection (kill aura, reach)
  • Movement analysis
  • Behavioral profiling

What's the flag-only mode?

When ml.flag-only: true in config, the plugin will only notify staff about suspicious players without applying automatic punishments. This is useful for testing or servers that want manual punishment.

What are punishment levels?

6 escalating levels:

  1. Warning message
  2. Mining debuff (slow mining)
  3. Inventory lock
  4. Area restriction
  5. Temporary ban
  6. Permanent ban

Troubleshooting

Players are getting false positives

  1. Reduce ML sensitivity in config
  2. Enable flag-only mode for testing
  3. Check if players have legitimate high efficiency tools
  4. Review suspicious threshold settings

Performance issues

  1. Reduce max players in auto-analysis
  2. Increase analysis intervals
  3. Disable unused features
  4. Ensure you have enough RAM

Commands not working

  1. Check permissions
  2. Verify command-hiding is disabled or you have owml.use
  3. Try /owml reload

Decoy ores not appearing

  1. Check decoy.enabled in config
  2. Verify ore threshold is not too high
  3. Ensure requireBuried is achievable in your world

ML System

How do I train the ML system?

Use the bot training system:

/owml ml spawn XRAY_CHEATER 5
/owml ml spawn NORMAL_MINER 5

Then analyze players:

/owml ml analyze <player>

Can I use my own data?

Yes, use the train command:

/owml ml train <player> cheater
/owml ml train <player> normal

What's the difference between bot types?

  • NORMAL_MINER - Legitimate mining patterns
  • XRAY_CHEATER - Cheating patterns to train detection
  • Other types - Various mining behaviors

Discord

How do I set up Discord?

  1. Create a Discord Bot at https://discord.com/developers/applications
  2. Enable Message Content Intent
  3. Add bot to your server
  4. Configure in config.yml:
    • bot-token
    • notification-channel
    • command-channel

What alerts can I receive?

  • X-ray detection alerts
  • Suspicious mining alerts
  • Punishment applied alerts
  • Staff action logs
  • Appeal updates

Support

Where can I get help?

  • SpigotMC discussion thread
  • GitHub Issues
  • Discord support server

How do I report bugs?

Open an issue on GitHub with:

  • Server version
  • Plugin version
  • Error logs
  • Steps to reproduce

Clone this wiki locally