Polyscrap

Installation

bun install

Usage

bun find-traders

Trader Scoring Algorithm

The scoring system evaluates traders based on 5 weighted factors, producing a composite score from 0-100.

Default Weights

Factor	Weight	Description
ROI	30%	Return on Investment
Win Rate	25%	Percentage of profitable trades
Consistency	20%	Stability of returns
Volume	15%	Total trading volume
Recent Activity	10%	How recently the trader was active

Metrics Calculation

ROI (Return on Investment)

ROI = (Total Realized PnL / Total Invested) × 100

Measures overall profitability relative to capital deployed.

Win Rate

Win Rate = (Winning Positions / Total Positions) × 100

A position is "winning" if realizedPnl > 0.

Consistency Score

Measures the stability of per-trade returns using variance:

1. Calculate per-position ROI: positionROI = realizedPnl / totalBought
2. Compute standard deviation of all position ROIs
3. Transform to 0-1 scale: consistency = 1 / (1 + stdDev)

Higher consistency (closer to 1) means more predictable returns.

Volume Score

Uses logarithmic scaling to prevent large traders from dominating:

volumeScore = normalize(log(totalVolume + 1))

Recency Score

Linear decay based on days since last activity:

• Active within 1 day: 100
• Active 30+ days ago: 0
• Between: Linear interpolation

recencyScore = max(0, 100 - (daysSinceActive / 30) × 100)

Normalization

ROI, Win Rate, and Volume scores are normalized using min-max normalization across all analyzed traders:

normalizedScore = ((value - min) / (max - min)) × 100

This ensures scores are relative to the current trader pool.

Composite Score

The final score is a weighted sum:

compositeScore = (roiScore × 0.30) +
                 (winRateScore × 0.25) +
                 (consistencyScore × 0.20) +
                 (volumeScore × 0.15) +
                 (recencyScore × 0.10)

Filtering Criteria

Traders are excluded if they don't meet minimum requirements:

• Minimum trades: 10 closed positions
• Minimum volume: $100 total invested