The 11-Factor Signal Engine
Eleven independent factors. Each scored separately. Normalized across 185 max points into a single conviction score and market bias. Built from 9+ years of live trading data — now fully automated.
Price Momentum
The highest-weighted factor. ADX-adjusted so trending markets get full weight while ranging markets are discounted before a signal is scored. A level-break bonus of up to +10 points fires when price breaks a key structure level with momentum behind it — the highest-confidence entry setup in the engine.
- ▸ADX strength (regime-adjusts the base score)
- ▸Directional momentum magnitude and consistency
- ▸Key structure level proximity and breakout confirmation
- ▸Level-break bonus: +10 pts on high-conviction structural breaks
- BTC ADX 32, breaking major resistance on momentum — BULLISH +35
- ETH ADX 14 (ranging) — momentum score discounted to +15
- SOL failed breakout, momentum reversed below level — BEARISH
Technical Analysis
A composite score pulling from multi-timeframe TA — RSI, MACD, EMA stack alignment, and candlestick structure — combined with TD Sequential timing signals. Like Price Momentum, the score is ADX regime-adjusted: trending markets produce cleaner TA reads than choppy ones, and the engine knows the difference.
- ▸RSI momentum state and divergence detection
- ▸MACD histogram direction and histogram expansion
- ▸EMA stack alignment (short vs mid vs long-term trend)
- ▸TD Sequential price count for timing precision
- BTC: EMA stack bullish, RSI reclaiming 50, MACD crossing — BULLISH +28
- ETH: RSI bearish divergence at resistance, TD 9 count — reversal signal
- SOL: EMA stack inverted, MACD below zero — BEARISH confirmation
InDecision Bias
The proprietary Knox formula — the intellectual core of the engine. Born from options market analysis and adapted natively for perpetual futures. Weights crowd sentiment against actual price action to detect divergence between what the crowd believes and what price is doing. When these diverge, alpha follows.
- ▸Perpetual funding rate as a sentiment proxy
- ▸Price action vs crowd positioning divergence
- ▸Sentiment extremes (fear-driven vs greed-driven setups)
- ▸Smart money positioning relative to retail flow
- Funding deeply negative but price holding key support — BULLISH bias, +20
- Funding at 0.08% (extreme greed) while price stalling — BEARISH bias
- Neutral funding, no sentiment signal — InDecision Bias abstains
Funding Rate
The perp market's most honest real-time signal. Extreme positive funding means the market is overloaded with longs and crowded in one direction — a setup for a violent reversion. The engine applies an adverse funding penalty: if open interest is running against your directional signal, the entire score is discounted proportionally.
- ▸Funding rate magnitude and direction
- ▸Funding vs price divergence (contrarian setup detector)
- ▸Cross-exchange funding consistency
- ▸Adverse funding penalty when funding opposes trade direction
- BTC funding at -0.05% (extreme fear shorts) — BULLISH setup +20
- ETH funding at +0.07% (crowded longs), price weakening — caution
- SOL funding neutral — no penalty, no premium applied
ATR Expansion
The ATR regime filter measures whether the market is in expansion or compression relative to its own recent history. Expanding ATR vs the rolling median signals a trending move with real momentum. Compressing ATR signals chop, consolidation, or a pending reversal. The engine uses this to detect regime shifts before they become obvious on a price chart.
- ▸Current ATR vs rolling median (expansion ratio)
- ▸ATR trend direction (expanding vs contracting)
- ▸Regime classification (trending vs ranging)
- ▸Velocity of ATR change (acceleration detection)
- BTC ATR at 1.8× median — trending regime confirmed, +15 pts
- ETH ATR at 0.6× median — ranging/compression, ATR factor discounted
- SOL ATR spiking from compression — breakout regime beginning
Open Interest
Open interest tells you whether the market's participation is expanding or contracting. Rising price with rising OI = real trend with new money coming in. Rising price with falling OI = short-covering rally — no new buyers, just trapped shorts unwinding. The engine distinguishes between these and scores accordingly.
- ▸OI trend direction (expanding vs contracting)
- ▸OI change aligned with price direction (confirms vs contradicts)
- ▸Cross-exchange OI consistency
- ▸OI spike detection (sudden institutional entry)
- BTC price up, OI up — real trend with new participation, +15
- ETH price up, OI down — short squeeze, not a real trend, reduced score
- SOL OI surging into price drop — new shorts entering, BEARISH confirmation
Volume Confirmation
A dynamic rolling baseline that adapts to each asset's normal volume profile — not a static average. Price moves on low relative volume get penalized because they lack participation. Breakouts without volume are suspect. Volume is the engine's lie detector for price action.
- ▸Relative volume vs dynamic rolling baseline
- ▸Volume trend direction (accumulation vs distribution pressure)
- ▸Price-volume divergence (price up, volume down = warning)
- ▸Breakout volume confirmation (volume must lead price)
- BTC breakout on 2.4× avg volume — participation confirmed, +10
- ETH rally on 0.3× avg volume — low conviction, volume score discounted
- SOL volume surging into downtrend — distribution confirmed, BEARISH
Trend Alignment
The only factor in the engine that can subtract from the total score. Multi-timeframe consensus measures whether the 15m, 1H, 4H, and Daily are all pointing the same direction. Full alignment amplifies every other factor. Conflicting timeframes introduce a penalty — the signal is noise, not edge. A bullish 1H inside a bearish daily trend is a trap, not a setup.
- ▸15-minute trend direction and momentum
- ▸1H and 4H intermediate trend alignment
- ▸Daily timeframe trend dominance
- ▸Conflict penalty: opposing timeframes subtract from conviction
- BTC: all 4 timeframes bullish — full alignment, +10 pts
- ETH: 1H bullish, 4H neutral, Daily bearish — conflict, -5 penalty
- SOL: all timeframes bearish — max BEARISH alignment, +10 bear side
VWAP Alignment
Volume-Weighted Average Price is the institutional benchmark — the level most algorithmic strategies anchor to. Price above VWAP signals buy-side dominance for the session. Price below VWAP signals sell-side control. A bullish signal where price is trading above VWAP has more institutional support behind it than the same signal below VWAP.
- ▸Price position relative to session VWAP
- ▸VWAP slope (rising vs flat vs declining)
- ▸Distance from VWAP (extension risk detection)
- ▸VWAP reclaim vs rejection (key intraday signal)
- BTC trading above VWAP, VWAP rising — buy-side dominant, +10
- ETH below VWAP, rejected reclaim attempt — sell-side control, bearish add
- SOL far above VWAP (overextended) — extension risk flags, score capped
Pattern Score
The chart pattern engine scores classical formations — pennants, flags, breakouts, wedges — but only when they form at key levels and are confirmed by volume. Random patterns in empty space score zero. Patterns at significant structure score full points. TD Sequential adds timing precision: a 9-count at a key level during a pattern breakout is one of the highest-accuracy confluences the engine can register.
- ▸Pattern identification (pennant, flag, triangle, wedge, range)
- ▸Key level context (pattern at support/resistance scores higher)
- ▸Volume confirmation of pattern breakout
- ▸TD Sequential price count for entry timing
- BTC bull flag at key support, breaking out on volume — +10
- ETH pennant forming, no key level context — partial score only
- SOL TD 9 count at major resistance, bearish engulfing — reversal signal
Liquidity Sweep
Stop-hunt detection. When price wicks below a significant low and immediately reclaims the level, it has swept the stop-loss orders clustered beneath it — trapping shorts and clearing the path for a reversal. This fingerprint is left by smart money moving size through stop-dense zones. One of the highest-accuracy setups in the engine, and one of the most misunderstood by retail traders.
- ▸Wick magnitude below key low (sweep detection)
- ▸Immediate price reclaim speed (quality of the sweep)
- ▸Volume during the sweep (smart money signature)
- ▸Prior level significance (higher significance = larger liquidity pool)
- BTC wicks below key low, reclaims in same candle on volume — BULLISH sweep, +10
- ETH wick below support, no reclaim — breakdown, not a sweep
- SOL long wick below major low, closes above — classic liquidity grab, BULLISH
When the Model Speaks — and When It Doesn't
InDecision does not force a call when the data is unclear. A forced call under uncertainty is worse than no call. No signal is also a signal.
Why 11 Factors?
Most traders use 1-2 indicators. They get confirmation bias and overfit to recent market conditions. The InDecision Engine was designed around a core principle: no single factor is reliable enough on its own.
The framework started as a 6-factor manual model. As it evolved into an automated signal engine, the factors expanded to 11 — each one filling a gap the others couldn't cover. Price Momentum and Technical Analysis handle price structure. InDecision Bias and Funding Rate handle crowd positioning and perp market dynamics. ATR, OI, and Volume handle regime and participation. Trend Alignment, VWAP, Pattern Score, and Liquidity Sweep handle confluence, timing, and institutional fingerprints.
The score weights are not arbitrary. Price Momentum earns the most points (35) because directional momentum with ADX confirmation is the single most predictive signal in trending crypto markets. Technical Analysis follows (30) because TA confluence — when RSI, MACD, and EMA stack all agree — is the second most reliable filter. InDecision Bias and Funding Rate each cap at 20 because perp market dynamics are the defining edge in crypto over equity markets.
Trend Alignment is the only factor that can subtract from the score by design. A signal that conflicts across timeframes isn't just weak — it's a trap. The engine penalizes timeframe conflict instead of ignoring it, which is the difference between a systematic framework and a wishful one.
Newsletter Launching April 2nd
Weekly signals with full 11-factor breakdowns. Every call explained. No black box. Join the waitlist — first issue drops April 2nd.