GodzillaBTCGodzillaBTC Methodology
GodzillaBTC is designed as a multi-factor research framework rather than a single-indicator strategy.
Signal pipeline
Market data → feature engine → regime detection → multiple model families → probability calibration → ensemble decision → risk engine → signal history → outcome evaluation.
Model families
- Trend: moving averages, ADX, normalized slope and higher-timeframe structure.
- Momentum: RSI, MACD, ROC and stochastic features.
- Volatility: ATR, realized volatility and Bollinger conditions.
- Statistics: returns, z-scores, distributions and rolling relationships.
- Structure: swing points, breakouts and support/resistance.
- Risk: expected value, fractional-Kelly-inspired sizing and drawdown controls.
Probability definition
A useful probability must define an event. For example: “estimated probability that price reaches +1R before −1R within the next N candles.” Calibration and out-of-sample testing are required before treating a probability as reliable.