ChartGuru Gem Finder Review: How Results Are Scored
How ChartGuru Gem Finder scores stocks and crypto—data-support vs confidence, classifications, risk labels, and what the product does not claim.
If you're evaluating whether to trust Gem Finder output at all, the fair question isn't "does it work" — no screening tool can promise that, and any that claims to should be treated with suspicion. The fair question is: what is the score actually based on, and does the tool show its work?
What generates a Gem Finder score
Every ranked result is scored from live, usable inputs — not a static marketing number. On both stocks and crypto, that includes market confirmation (volume, momentum, persistence), liquidity quality, and opportunity / size context. Stocks add fundamental enrichment (growth, profitability, valuation, catalysts) and filing-risk gates where available. Crypto adds tokenomics, on-chain growth signals, and minimum contract-safety checks.
The published data-support score reflects how strongly those available inputs support further research. Missing inputs are not backfilled with placeholders; incompleteness reduces confidence or forces an Insufficient data classification instead of inventing a score.
Confidence is not "how bullish the chart is"
This is the distinction most gem-tool marketing blurs:
| Field | Meaning |
|---|---|
| Data-support score | Strength of the ranking from available features |
| Confidence | Quality of those inputs (completeness, freshness, agreement, persistence) |
| Classification | Label for data-support quality (e.g. high / strong / moderate / limited / low / insufficient) |
| Risk | Independent risk label shown beside the score |
A moderate data-support result with honest confidence is more useful than a "high" call with no way to check the inputs. Gem Finder shows weaker classifications on purpose — filtering them out would make the board look cleaner than the data supports.
What a result includes in practice
A Gem Finder result typically includes:
- The asset (stock ticker or crypto identity)
- Data-support score and confidence
- Classification and risk label
- Safety or filing context where applicable
- Strengths and risks derived from the strongest and weakest features
None of this is presented as advice to act on. It's a structured research ranking — a starting point for your own technical and fundamental review, not a signal to follow blindly.
What Gem Finder doesn't claim
It doesn't claim a historical win rate or backtested accuracy percentage unless that figure is generated live from actual tracked results — hardcoded accuracy claims are against how ChartGuru builds this product. If you see an accuracy stat in marketing that doesn't match live data, treat that as a copy bug, not the product standard.
It also does not emit investment-merit labels ("buy," "pick," "opportunity"). Classifications describe data support only.
The honest limitation
Gem Finder screens for multi-factor data support and risk. You still own liquidity depth checks at your size, deeper contract diligence on crypto low-caps, and company-specific news risk on small-cap stocks. For chart bias, key levels, and invalidation, open full technical analysis on the name — those fields belong to the chart read, not the discovery board.
What Gem Finder is built to solve is the research starting point: surfacing ranked names worth that closer look, with the reasoning shown, instead of a tip with no way to check the logic behind it.
FAQ
Why are some scores null?
When data completeness or safety / filing gates fail, numeric scores stay null and classification becomes Insufficient data — by design.
Is confidence the same as win rate?
No. Confidence measures input quality for the data-support score, not historical hit rate.