🎯 Accuracy Scorecards
Full transparency on how well GripNews predictions perform. Every signal, forecast, and trust weight is tracked and scored.
🔒 Source Trust Weights
Each source type earns trust over time based on how often its signals lead to accurate outcomes. These weights directly influence confidence scoring.
| Source Type | Trust Weight | Visual | Reliability |
|---|---|---|---|
| Patch Notes | 0.00 | ⚠️ Unverified | |
| Steam News | 0.00 | ⚠️ Unverified | |
| Developer Blog | 0.00 | ⚠️ Unverified | |
| Press Release | 0.00 | ⚠️ Unverified | |
| Driver Pattern | 0.00 | ⚠️ Unverified | |
| Esports Event | 0.00 | ⚠️ Unverified | |
| Regression | 0.00 | ⚠️ Unverified | |
| Frequency Spike | 0.00 | ⚠️ Unverified | |
| Steam Community | 0.00 | ⚠️ Unverified | |
| Hardware Correlation | 0.00 | ⚠️ Unverified | |
| Emerging Issue | 0.00 | ⚠️ Unverified | |
| Youtube Coverage | 0.00 | ⚠️ Unverified | |
| Cross Game | 0.00 | ⚠️ Unverified | |
| Reddit Trend | 0.00 | ⚠️ Unverified | |
| Anomaly | 0.00 | ⚠️ Unverified | |
| Social Spike | 0.00 | ⚠️ Unverified | |
| Rumour | 0.00 | ⚠️ Unverified |
📋 Forecast Outcomes
Every prediction GripNews makes is tracked. When the prediction window closes, we check the outcome and update our accuracy score.
Forecasts are being tracked. Results will appear here as prediction windows close.
The system generates forecasts based on momentum patterns and signal intelligence.
📐 How We Score Accuracy
📡 Signal Trust Weights
Each data source (Steam news, patch notes, social media, leaks, etc.) starts with a baseline trust weight. Over time, the system adjusts these weights based on how often signals from each source lead to confirmed outcomes. High-trust sources influence momentum scores and predictions more heavily.
🔮 Forecast Tracking
When the system generates a forecast (e.g., "Fortnite momentum will rise in the next 7 days"), it records the current momentum score and prediction details. When the window closes, the outcome checker compares the predicted direction against what actually happened. The result (correct/incorrect/expired) feeds back into the accuracy score.
🔄 Self-Improving Loop
Every day at 07:30 UTC, the outcome checker runs: it snapshots momentum history, checks pending forecasts, scores signal accuracy, updates trust weights, and advances trend lifecycles. This creates a feedback loop where the system gets smarter over time — unreliable sources lose influence, and accurate patterns get amplified.
📊 Transparency
Everything on this page is pulled live from the Intelligence API. No cherry-picking, no hiding bad predictions. When we get it wrong, you'll see it here. That's the point — transparency builds trust.