Beneat

Layer 3: Predictive Zones #

Predictive zones are generated when our ML models detect multiple high-accuracy actors converging at similar price levels—identifying where price is likely to go before it arrives.

Our deep learning models analyze wallet tracking data to detect meaningful convergence patterns. When several high-scoring wallets independently accumulate at comparable price levels, the models generate predictive zones with associated confidence scores.

Why This Works

A single high-accuracy actor buying at a price level could be coincidence. Five independent actors converging at the same zone is predictive signal worth attention.

Zone Generation #

ML Detection Process #

Our deep learning models continuously analyze wallet tracking data to identify predictive zones:

┌──────────────────────────────────────────────────────────────┐
│              WALLET TRACKING LAYER                           │
│      Continuous monitoring of high-accuracy actors           │
└────────────────────────┬─────────────────────────────────────┘
                         │
                         ▼
┌──────────────────────────────────────────────────────────────┐
│              ML PATTERN DETECTION                            │
│      Deep learning identifies convergence patterns           │
└────────────────────────┬─────────────────────────────────────┘
                         │
                         ▼
┌──────────────────────────────────────────────────────────────┐
│              PREDICTIVE SCORING                              │
│      Models assign probability and confidence                │
└────────────────────────┬─────────────────────────────────────┘
                         │
                         ▼
┌──────────────────────────────────────────────────────────────┐
│              ZONE OUTPUT                                     │
│              High • Medium • Low Confidence                  │
└──────────────────────────────────────────────────────────────┘

Zone Confidence Tiers #

Zone Confidence	Description
High	Multiple elite actors with high conviction, very high probability
Medium	Several quality actors showing clear interest, solid probability
Low	Early zone from qualified actors, moderate probability

Dynamic Thresholds

Our zone thresholds adapt based on market conditions and are calibrated through continuous backtesting. The specific parameters are proprietary.

Zone Components #

Each probability zone includes comprehensive data to inform your decisions:

Token Information: Contract address, symbol, chain, and liquidity data
Zone Confidence: Classification based on our proprietary criteria
Price Range: Upper and lower bounds of the probability zone
Aggregated Metrics: Actor participation, accumulation patterns, and timing
Probability Score: Statistical measure of zone reliability
Timestamp: Precise timing of zone generation

// Example zone structure (simplified)
interface ProbabilityZone {
  token: TokenInfo;
  range: { lower: number; upper: number };
  confidence: "high" | "medium" | "low";
  probability: number;
  direction: "accumulation" | "distribution";
  metrics: ZoneMetrics;
  timestamp: string;
}

Predictive Performance #

Our zones have demonstrated consistent predictive value across market conditions. Based on backtesting across multiple market cycles:

Zone Confidence	Historical Hit Rate	Notes
High	70-80% range	Highest conviction, lowest frequency
Medium	60-70% range	Balanced quality and frequency
Low	50-60% range	Early indicators, higher variance

These ranges represent aggregate performance across all tracked tokens and chains. Individual results vary based on:

Market conditions (bull vs bear vs sideways)
Token liquidity and market cap
Entry and exit timing
Position sizing

Important Disclaimers

Past performance does not guarantee future results. These ranges are based on historical backtesting and may not reflect future performance. Detailed, up-to-date metrics are available via API. Always conduct your own research.