Score Composition
The score combines four dimensions. The two core dimensions establish the baseline; the two modifier dimensions adjust it up or down based on depth and risk profile.Weights are applied deterministically using a rule-based engine. No machine learning — every score is explainable.
Score Ranges
Score on Wallet Pages
On any wallet page you will see:- The score badge, color-coded by range (High Confidence, Solid, Mixed, or Low Confidence)
- The classification tag (Early Mover, Momentum Trader, and so on)
- A confidence indicator (Low / Medium / High) based on total qualifying event count
- A breakdown section showing each dimension’s individual contribution to the final score
Score on Token Pages
On token pages, Cortex aggregates reputation data across all top holders and surfaces the most actionable summary inline — for example:3 of top 10 holders are classified Early Mover (score >70), currently net accumulating.This gives you an immediately actionable signal without requiring you to click into each wallet individually. You can assess the quality of a token’s holder base in seconds, directly from the token page.
Caveats
- Scores reflect on-chain activity only — off-chain behavior, social signals, and external context are not captured
- A new wallet with no history scores 0 by default, which means unrated, not bad — do not interpret 0 as a negative signal for wallets with no recorded events
- Scores can decrease if recent behavior diverges from historical patterns — a formerly high-scoring wallet that begins showing poor timing or high drawdowns will see its score adjust downward over time