Alpha++
Amplified edge on top of ensemble v1: market breadth tilt, walk-forward quality gates, confluence bonus, and regime-adaptive thresholds. Model ensemble-v2.0 powers watchlist, forward log, and daily snapshots.
Regime
Risk-off
Defensive regime — longs need alpha ≥ 82; fewer assets pass the gates.
Breadth advancing
19%
Basket α vs BTC
—
Top picks
0
Alpha watchlist
Ranked by alpha score (signal + historical edge + confluence). Classic signal view · Forward log
No assets pass Alpha++ gates in Risk-off (effective min score ~78). This is expected when breadth is neutral and liquidity / walk-forward quality filters are strict — not an error.
Near-misses — highest alpha scores that did not qualify:
| Asset | Alpha | Dir | Min needed | Gap |
|---|---|---|---|---|
| BEAT | 63 D | DOWN | 78 | +15 pts |
| ADI | 60 D | UP | 78 | +18 pts |
| PUMP | 55 D | UP | 78 | +23 pts |
| HASH | 52 D | UP | 78 | +26 pts |
| LTC | 51 D | UP | 78 | +27 pts |
| ALGO | 50 D | UP | 78 | +28 pts |
| OKB | 49 F | UP | 78 | +29 pts |
| CC | 49 F | UP | 78 | +29 pts |
Alpha score history
Daily snapshots from the worker — one point per UTC day per asset. Tracks how Alpha++ evolved over time. Export CSV ↓
Need at least 2 daily snapshots to plot.
Alpha++ formula
base_vote = ensemble_v1 (linreg+momentum+onchain+forecast) amplified = 85%·base + 15%·breadth_tilt × quality_mult (backtest accuracy, vol, liquidity) × confluence_bonus (≥3 agreeing sub-signals) alpha_score = blend(composite, accuracy, strategy_α, win_rate) × confidence min_score: risk_on 68 · chop 72 · risk_off 78 (longs need ≥82 in risk_off)