Strix EDGE Whitepaper

Table of Contents

1. The Problem with Single-System Trading

The overwhelming majority of retail trading tools offer a single analytical lens: an RSI scanner, an MACD crossover alert, a chart pattern detector. Each of these tools, in isolation, produces a signal-to-noise ratio that makes consistent profitability extremely difficult. An RSI reading of 30 on a 15-minute chart tells you almost nothing without context from higher timeframes, volume confirmation, structural analysis, and market-wide positioning data.

The core insight behind Strix EDGE is that no single analytical system is reliable in isolation, but the probability of a correct directional call increases dramatically when multiple independent systems — built on fundamentally different mathematical and structural foundations — converge on the same conclusion.

If a momentum oscillator says "oversold," that's information. If simultaneously the market structure shows a Change of Character, the volume profile indicates institutional accumulation, an Elliott Wave count places price at the end of Wave 5, and Wyckoff analysis identifies a Spring event — that convergence of five independent methodologies pointing in the same direction is qualitatively different from any single indicator reading.

2. Platform Architecture

Strix EDGE operates as a client-side intelligence layer connected to ByBit's public market data infrastructure via REST API and WebSocket streams. The platform scans both Spot and Perpetual Futures markets (1,100+ USDT pairs) across four timeframes simultaneously: 15-minute, 1-hour, 4-hour, and Daily.

The 15 engines are deliberately organized into two categories based on their analytical methodology. Indicator Systems read numerical values from mathematical oscillators and classify them. Scanner Systems detect geometric patterns, structural formations, and market phases. This separation ensures that the two confluence engines combine genuinely independent analytical schools, not variations of the same calculation.

3. Indicator Systems

Indicator systems read oscillator values, volume metrics, and funding rate data. They answer the question: "What do the numbers say right now?"

3.1 MTF Momentum Scanner

Calculates six momentum oscillators (RSI, Stochastic RSI, Stochastic, CCI, Williams %R, MFI) across all four timeframes. A signal is generated when 5 or 6 oscillators agree on overbought or oversold conditions across 3+ timeframes. This multi-oscillator consensus eliminates the false signals that plague any single indicator.

3.2 OBV Divergence + RVOL

Detects divergences between price action and On-Balance Volume, filtered by Relative Volume (RVOL). When price makes a new low but OBV makes a higher low — while volume is 1.5× or more above average — it signals institutional accumulation that hasn't yet been reflected in price. Requires confirmation on 3+ timeframes.

3.3 Divergence Scanner (6 Oscillators)

Extends divergence detection beyond OBV to six independent oscillators: MACD, RSI, Stochastic, CCI, Momentum (Rate of Change), and MFI. Uses TF-adaptive parameters for pivot detection (depth, distance, amplitude thresholds calibrated per timeframe). A signal requires 4+ of 6 oscillators showing divergence — a threshold that eliminates virtually all false positives.

3.4 Funding Rate & Open Interest

Analyzes perpetual futures funding rates to detect squeeze conditions. A Long Squeeze (negative funding + falling price) indicates shorts are paying while price drops — creating conditions for a violent short-covering rally. A Short Squeeze (positive funding + rising price) signals the opposite. Also flags overextended positioning when annualized funding exceeds ±36.5%.

3.5 Advanced Flow Scanner

Combines three independent indicator systems in a single efficient scan: Ichimoku Cloud (cloud position, Tenkan/Kijun cross, Kumo twist — 0-3 points), Bollinger/Keltner Squeeze Momentum (squeeze state + momentum direction — 0-2 points), and Weis Wave Volume (up-wave vs down-wave volume ratio — 0-2 points). Combined score 0-7; signals generated at 5+.

4. Scanner Systems

Scanner systems detect patterns, structures, and formations. They answer the question: "What is the market doing structurally?"

4.1 Trend Scanner

Evaluates trend strength using SMA(50), SMA(200), RSI(14), 24-hour momentum, and relative volume. Produces a weighted confluence score (0-100) combining MA position (50pts), MA alignment (15pts), RSI strength (20pts), momentum (15pts), and volume bonus (0-20pts). Detects Golden Cross and Death Cross events. Signal threshold: score ≥ 85 with 3+ TF agreement.

4.2 Market Structure Scanner

Implements Smart Money Concepts: classifies market structure as HH/HL (uptrend) or LH/LL (downtrend) using TF-adaptive pivot detection. Detects Break of Structure (BOS — continuation) and Change of Character (CHoCH — reversal). CHoCH on 3+ timeframes with trend alignment generates a signal — one of the highest-probability reversal detections available.

4.3 Liquidity Heatmap

Scans order book depth for all USDT Perpetual Futures. Detects bid/ask walls (orders 3× above average), calculates imbalance ratio (bid volume ÷ ask volume within 2% of price), and estimates liquidation levels. Signals generated only at imbalance ratios exceeding 3:1 — indicating genuine institutional order flow, not retail noise.

4.4 Price Action Scanner

Detects 14 candlestick patterns (Engulfing, Morning/Evening Star, Three Soldiers/Crows, Hammer, Shooting Star, Doji variants, Marubozu, Piercing Line, Dark Cloud Cover) with Support/Resistance zone confluence. S/R zones are computed using pivot high/low clustering with 0.5% tolerance. A pattern alone is not a signal — S/R confluence is required. Score = pattern strength × 10 + S/R proximity bonus. Signals require patterns on 4/4 TF with score ≥ 40.

4.5 Chart Pattern Scanner

Six geometric pattern detectors with reliability scoring (0-100): Head & Shoulders (+ Inverse), Double Top/Bottom, Ascending/Descending/Symmetrical Triangles, Rising/Falling Wedges, Bull/Bear Flags, and Cup & Handle. Each detector evaluates multiple quality metrics (symmetry, spacing, volume decline, breakdown confirmation). Signals require the same pattern type detected on 4/4 timeframes with reliability ≥ 70.

4.6 Fair Value Gap Scanner

Detects institutional imbalance zones where candle[i-2].high < candle[i].low (bullish FVG) or candle[i-2].low > candle[i].high (bearish FVG), filtered by ATR(14) × 0.3 minimum gap size and EMA(20) trend alignment. Tracks open vs filled status and distance from current price. Confluence score (0-100) weights distance (30pts), direction agreement (25pts), TF coverage (25pts), and gap count (20pts). Signal at score = 100.

4.7 Elliott Wave Scanner

Identifies impulse (5-wave) and corrective (ABC) wave structures using zigzag pivot detection with depth-5. Validates three Elliott rules for impulse waves: Wave 3 is never the shortest, Wave 2 retracement ≤ 61.8% of Wave 1, Wave 4 retracement ≤ 61.8% of Wave 3. Calculates Fibonacci extension targets (1.618, 2.618). Signals require the same wave pattern on 3+ TF with confidence ≥ 95.

4.8 Wyckoff Scanner

Detects the five phases of Wyckoff Accumulation (Selling Climax → Automatic Rally → Secondary Test → Spring → Sign of Strength) and Distribution (Buying Climax → Automatic Reaction → ST → UTAD → Sign of Weakness). Volume confirmation required: Climax events need volume ≥ 1.8× average, SOS/SOW need ≥ 1.3×. Confluence score (0-100) weights phases (30pts), Spring/UTAD (25pts), SOS/SOW (15pts), TF coverage (20pts), and confidence bonus (10pts). Signal at confluence ≥ 90.

5. The Confluence Framework

The confluence engines are the core innovation of Strix EDGE. They take the outputs of all individual systems and identify pairs where multiple independent methodologies agree on direction.

5.1 Indicator Confluence

Combines votes from 5 indicator systems (Momentum, OBV, Divergence 6-Osc, Advanced Flow, Funding) using relaxed individual thresholds. Each system casts a directional vote (LONG, SHORT, or neutral). Signal when 3+ of 5 agree. The relaxed individual thresholds are deliberate — it's the agreement count that provides the filtering, not the individual system's confidence.

5.2 Scanner Confluence

Combines votes from 7 scanner systems (Trend, Structure, Price Action, Chart Pattern, FVG, Elliott, Wyckoff). Liquidity Heatmap is excluded because it requires a separate API call (order book depth vs kline data). Signal when 6+ of 7 agree — an extremely high bar that eliminates virtually all noise. When six independent pattern detection systems built on completely different methodologies all point in the same direction for the same asset, the statistical significance is substantial.

6. Data Architecture & Real-Time Pipeline

All market data is sourced from ByBit's v5 public API infrastructure. The platform uses two data channels:

REST API — Kline (OHLCV candlestick) data for analytical computation, ticker data for market overview, and order book depth for liquidity analysis. Scan cycles run every 60-180 seconds depending on the tool. Results are cached client-side and restored instantly on navigation.

WebSocket — Real-time tick-by-tick price updates for BTC/USDT and watched assets. Sub-second latency via ByBit's streaming infrastructure (wss://stream.bybit.com/v5/public/spot).

Kline requests are batched (6 pairs per batch) to respect API rate limits while maintaining scan speed. A full market scan (837+ pairs × 4 timeframes) completes in 45-60 seconds. Results are rendered in real-time as each batch completes — traders see signals as they're discovered, not after the scan finishes.

7. Scoring, Filtering & Signal Quality

Every system employs strict filtering to minimize false signals. The philosophy: a missed signal is a minor cost; a false signal erodes trust and capital. Each scanner's thresholds have been calibrated so that from 837+ pairs scanned, typically 5-30 signals pass — not 200+.

8. Risk Considerations

Strix EDGE is an analytical tool, not a trading bot. No signal — regardless of how many systems agree — guarantees a profitable trade. Cryptocurrency markets are subject to exogenous shocks (regulatory actions, exchange failures, black swan events) that no technical analysis can anticipate.

The confluence framework increases probability, not certainty. Even with 7/7 scanner agreement, a trade can fail. Position sizing, stop-loss discipline, and portfolio risk management remain the trader's responsibility. The platform surfaces high-probability setups; the trader manages risk.

Elliott Wave analysis is inherently subjective — different analysts routinely disagree on wave counts. The automated detection uses strict rule validation and cross-timeframe confirmation to mitigate this, but no wave count should be treated as definitive. Similarly, Wyckoff phase detection on shorter timeframes (15m, 1H) is less reliable than on daily or weekly charts.

9. Conclusion

Strix EDGE represents a shift from single-indicator trading tools to multi-system intelligence. By running 16 independent analytical engines across 1,100+ pairs and surfacing only the setups where multiple independent methodologies agree, the platform provides a level of analytical depth that was previously available only to institutional trading desks with teams of specialized analysts.

The architecture is deliberately modular — new analytical engines can be added to the confluence framework without disrupting existing systems. Each engine operates independently, casts its vote independently, and the confluence count provides the final filter. This design ensures that the platform grows more accurate as more independent systems are added, not more noisy.