BTC 1h move vs ±1.5% threshold — frozen at decision time by scanner
[5] OI × CVD quadrants (graded MTF signals)
—
CVD-sell = takers aggressing in signal direction. Study cells: fresh-short (OIfix+sell) and new-long (OIup+buy) positive; distribution-long (OIup+sell) is the trap.
[4] SHORT signals by 30m OI change
—
Blocker candidate: shorting after a deleveraging flush (OI collapsing).
During basket events (any allowlist symbol |1h|≥2%, armed 120min), ceiling-band LONGs are logged as would-trades — still blocked live. Backtest: 7/7, +7R. Go-live decision after several events confirm.
[3] 1h EMA-blocked signals (CHoCH companion)
—
—
Other shadows
—
Full detail: python3 shadow_report.py · decisions after 3–4 weeks out-of-sample, one variable at a time.
1h-regime pullback-reclaim · SL 1.6×ATR15 · TP 2R · 72h. Sweet spot: ER48 0.15–0.25 (steady grind). ER<0.15 chop and ER≥0.40 parabolic lose; ATR-pctl≥60 helps. Cron every 5m, triggers only.
[7] Signed extension — exhaustion vs shallow entries
—
ext = price distance from 15m EMA21 in ATRs at signal. Study cells: SHORT shallow (ext −1..0) 70%/+0.41R gem; SHORT exhaustion trap (ext≤−2 & 4+ consecutive closes) ~0.00R; LONG chase (ext≥3) fine.
Pump = |1h move|≥20% on any liquid perp (24h vol≥$5M, |24h|≥25% prefilter). Motivation: AKE +449% on Sep 3 was invisible to the BTC-only event flag. Cells grade system signals fired inside pump windows.
Flash spikes (wick ≥1.4× / ≤0.7× prior close)
—
"Rejected" = spike bar closing back through the prior close — the blow-off signature. Post-spike dump stats: python3 shadow_report.py
Real-time Delta + Binance analysis. Also works from Telegram: /scan BTCUSD
Signal History — Day by Day (IST)
FromTo
📅
No history yet.
Silent Skip Log
🔇
No skips yet.
TradingView Webhook Setup (Essential Plan)
Alerts from TradingView fire to your scanner, which cross-confirms on Binance and sends Telegram with SL/TP automatically.