Anti-Cheat Overview¶
The Preads Platform incorporates an enterprise-grade Anti-Cheat engine designed to protect your network from fraudulent traffic, incentivize quality conversions, and maintain long-term advertiser trust through real-time behavioral and infrastructure auditing.
🛡️ Multi-Layered Defense Architecture¶
The system operates on a weighted risk-scoring model that aggregates signals from three distinct detection layers to calculate a consolidated Risk Score (0-100) for every click and conversion.
1. Infrastructure Layer (40% weight)¶
Uses real-time IP intelligence to identify high-risk network signatures:
- VPN/Proxy Detection: Real-time identification of tunnelled connections.
- TOR Exit Nodes: Automated blocking of anonymous browsing networks.
- Datacenter Traffic: Identifying non-residential IP ranges used for automated bot traffic.
2. Identity Layer (35% weight)¶
Leverages advanced device fingerprinting to bind physical hardware to user accounts:
- Multi-Accounting: Detection of users attempting to bypass limits using multiple uids on the same hardware.
- Bot Fingerprinting: Identification of headless browsers and automated script signatures.
3. Behavioral Layer (25% weight)¶
Analyzes user interaction patterns to identify non-human conversion activity:
- Fast Completion: Detection of conversions completed faster than humanly possible (e.g., watching a 30s video in 2s).
- Rapid Conversions: Identifies "burst" activity where a single user completes dozens of offers in a suspiciously short window.
- Suspicious Sequencing: Advanced pattern recognition for repetitive or non-linear navigation paths.
📊 Automated Decision Engine¶
The platform converts raw risk scores into automated administrative actions based on configurable thresholds.
| Risk Score | Risk Level | Automated Action |
|---|---|---|
| 0 - 19 | Low |
Allow: Immediate processing of clicks and conversions. |
| 20 - 39 | Medium |
Flag: Monitor user behavior without blocking traffic. |
| 40 - 59 | High |
Review: Queue conversion for manual administrative audit. |
| 60 - 100 | Critical |
Auto-Block: Immediate rejection of conversion and possible account suspension. |
[!IMPORTANT] TOR Exit Nodes are treated as critical infrastructure threats and are typically subject to immediate Auto-Blocking regardless of the weighted score.
🛠️ Management & Transparency¶
- Fraud Dashboard: Real-time visibility into flagged transactions, risk scores, and the specific "signals" (e.g.,
vpn_detected,fast_completion) that triggered the alert. - Audit Heritage: Every click and conversion stores a permanent snapshot of the fraud analysis at the time of the event, ensuring full transparency during publisher disputes.
- Custom Thresholds: Administrators can fine-tune "Threshold Sensitivity" (Critical, High, Medium) to match the risk profile of their specific network.