How Trust Score works

Trust Score combines three independent dimensions of trader quality into a single 0–100 number. A perfect 100 would mean: every signal is profitable, every win is much bigger than every loss, and the trader has posted enough signals that the result isn't luck.

In practice nobody scores 100. The current top scorers sit in the 70–80 range because the markets are noisy and crypto signals are roughly zero-sum. We think that's honest.

1. Win quality (40 points). The trader's average return per closed signal. We give the full 40 points only when the trader averages +3% or better on each call. Most traders cluster between 0% and +1%; the median active trader on our platform earns about +0.22% per call.
2. Risk discipline (30 points). The realised risk-to-reward ratio — the trader's average winning trade divided by their average losing trade. We give the full 30 points when this ratio is 3.0 or higher. We cap the input at 5 to prevent a single moonshot from pumping the score.
3. Consistency (30 points). The profit factor: total profit on winners divided by total loss on losers. A profit factor of 1 means break-even (zero points); we give the full 30 points at a profit factor of 2.2 or higher.
4. Sample-size dampening. The raw score from the three components above is then multiplied by n / (n + 30), where n is the number of closed signals. A trader with 10 closed signals only gets 25% credit; with 30 they get 50%; with 100 they get 77%; with 1,000 they get 97%. New traders cannot accidentally rank above veterans.

Each row is a live trader on the platform. The math runs the same for every account.

• Position sizing. We don't know how much capital each trader actually risked.
• Market regime. Bull markets flatter long-biased traders. We don't adjust for that yet.
• Multi-account / Sybil. The score reflects this account's public signals only. We don't deduplicate across alts.
• Drawdown. A trader with the same Trust Score might have had a brutal losing streak last month. We're working on a drawdown view.
• Slippage and fees. ROI is measured against signal price, not your actual fill.

The constants in the formula were calibrated from the empirical distribution of metrics across 67 active traders with at least 5 closed signals (snapshot 2026-04-25):

The thresholds are tuned so that roughly the top ~10% of active traders earn Verified, the next ~15% Solid, the middle ~40% Mixed, and the bottom ~25% Risky.

• v1.0·2026-04-25·Initial release. Components: 40/30/30. Bayesian shrinkage k=30.

We'll bump the version and document changes here whenever we tune the constants. Every score row in our database is tagged with the formula version it was computed under.