If you've used NavScope, you've seen the Safety Score sitting next to every token — a number between 0 and 10. Most people intuitively understand that higher is better. But what actually goes into that number?
This article breaks it down. No jargon. No hand-waving. Just the four components, what each one measures, and why each one matters to you as an investor.
Why a safety score at all?
Price aggregators give you prices. Market cap. Volume. Maybe a chart.
What they don't give you is any indication of whether the data you're looking at is reliable. A token with fabricated wash-trading volume looks identical on a standard chart to one with genuine organic volume. A token price-feed that's only tracked by a single low-liquidity exchange looks the same as one confirmed across five major venues.
The Safety Score exists to answer a different question than "what is this token worth?" It answers: "how much can I trust what I'm looking at?"
That's a prerequisite question. You need to answer it before any analysis is meaningful.
The four components
1. Data Quality (0–10 contribution)
This measures the cleanliness and completeness of the price feed data NavScope has collected for a token.
Specifically, it looks at:
- Gap frequency — how often are price updates missing entirely?
- Outlier density — how many data points fall outside statistically normal ranges?
- Feed consistency — do the values coming in make internal sense over time?
What this means for you: A token with a data quality problem is one where the chart you're seeing may not reflect what's actually happening on exchange. Sparse or erratic feeds are often a sign of very low genuine liquidity — or worse, a deliberately manipulated feed.
2. Source Count (0–10 contribution)
This is a simple measure of how many independent exchanges are reporting prices for a given token at any moment.
NavScope aggregates across 5 exchanges. A token reported by all 5 scores maximum on this component. A token only tracked by 1 scores near zero.
What this means for you: Cross-source confirmation is a basic credibility test. If a token's price is only visible on one exchange — especially a smaller one — there's no independent verification. You're trusting a single source. That's a risk factor regardless of what the chart looks like.
3. Volume Consistency (0–10 contribution)
This is where things get more sophisticated. Volume consistency measures whether the reported trading volume follows statistically plausible patterns over time.
Real trading volume has a signature: it rises and falls with market cycles, it correlates loosely with price movement, and it distributes reasonably across time windows. Wash trading and artificial volume inflation have a different signature: suspiciously round numbers, volume spikes with no corresponding price movement, perfectly uniform distribution across time.
The model is trained to distinguish between these patterns and score accordingly.
What this means for you: A high-volume token that scores poorly on volume consistency is a major red flag. It means the volume number you're seeing may not represent genuine market activity. Liquidity you can't trust is not liquidity.
4. Anomaly Penalty (deduction)
The first three components build the score up. The anomaly penalty knocks it down when the system detects specific warning signals:
- Sudden unexplained price discontinuities (e.g., a 40% gap between two consecutive data points)
- Data feed blackouts followed by resumption at a different price level
- Statistical signatures consistent with coordinated price manipulation
- Cross-source discrepancies that exceed normal arbitrage spreads
The penalty is applied in proportion to the severity and recency of anomalies detected. A single minor anomaly from six months ago has low impact. Repeated anomalies in the last 48 hours significantly depress the score.
What this means for you: The anomaly penalty is the component most likely to catch something actively wrong in real time. If a token's score drops suddenly, it's worth checking what triggered the penalty before acting on any price signal.
Three examples
TokenA — Score: 9.1/10
TokenA is a large-cap asset with deep liquidity. It's tracked across all 5 exchanges NavScope monitors. Its volume follows predictable daily patterns, correlates with price movement, and shows no statistical anomalies in the trailing 30 days. Data feeds update continuously with no gaps. This is what a trustworthy data profile looks like. A high score here doesn't mean "buy" — it means the information you're using to decide is reliable.
TokenB — Score: 5.4/10
TokenB is a mid-cap token tracked by 3 of 5 exchanges. Data quality is reasonable but there are occasional feed gaps. Volume is mostly consistent but shows some suspiciously uniform intraday patterns that trigger a mild volume consistency flag. No major anomalies, but the limited source coverage and volume irregularities drag the score to the middle. Proceed with additional research before acting on price signals.
TokenC — Score: 2.1/10
TokenC is only tracked by a single exchange. Feed data has significant gaps — roughly 30% of expected data points are missing. The reported 24-hour volume is orders of magnitude higher than comparable tokens in its category, with an artificially flat distribution across time windows that strongly suggests wash trading. Three price discontinuity anomalies were flagged in the last 7 days. A score this low doesn't automatically mean the project is fraudulent — but it does mean the data cannot be trusted, and any price chart you see for this token is built on an unreliable foundation.
What the Safety Score is not
It is not a fundamental analysis of the project. It does not evaluate the team, the whitepaper, the tokenomics, or whether the use case is credible. It does not predict whether the price will go up or down.
It is purely a measure of data integrity. Think of it as a confidence level on the information itself, not a verdict on the asset.
How to use it in practice
A practical threshold many users adopt: don't trade on any token scoring below 6.0 without significant additional due diligence. Scores below 4.0 warrant treating any price or volume data as unreliable by default.
Scores update continuously at navscope.io. You can monitor how a token's score changes over time — a declining score on a token you hold is a signal worth paying attention to.
NavScope is an independent data intelligence platform. Safety Scores are data quality indicators, not financial advice.