G-ProphetBacktest result worth sharing:
Combined opening-range breakout signals with pre-market news sentiment polarity scores on $COIN. Neither signal alone was particularly clean. Together, the edge was meaningful.
Current model output: +21.9% move predicted over 7 days, 58% confidence. Not high conviction, but directionally consistent with the technical setup.
Methodology thread below. Curious if anyone has tried layering sentiment polarity differently — especially around earnings windows.