Scores explained

What it does

Every score in CryptoBuddy is relative quality inside a class: how this gem, trade or spot buy compares with everything of its kind scanned at the time. It is not a probability of profit and not a prediction.

Scores are computed deterministically in code from provider data. The AI writes about the numbers; it never produces or adjusts one.

How it works

1. Provider data is normalised into evidence pillars; a field two providers carry counts once.
2. Pillars combine into a raw score, which is calibrated into the presented 0–100.
3. Security acts as a veto for gems, and unknown security caps a gem below green.
4. Data confidence (high, medium or limited) is worked out separately and shown beside the score, never hidden inside it.
5. Every stored score keeps its version, so results from different engine versions are never pooled silently.

Why it helps

• One vocabulary across the product: the same colour means the same thing everywhere.
• Confidence beside the score tells you how much evidence stands behind it.

What you’ll see

Report colours

Green 75–100 Strong · Light green 60–74 Worth a look · Amber 45–59 Speculative · Red below 45 Weak. Strong reads start at 75.

Score bands

75+ Strong · 60+ Worth a look · 45+ Speculative · below that Weak (and never in today’s top-rated).

Exceptional

Reserved for 90+: the rarest top-rated cards.

One scale everywhere

A coin you analyse yourself is scored on the same scale as today’s top-rated, against the same daily scale for its category: 90+ means one of today’s top 5 in that category.

Micro gems

Normal score tiers, always shown with Risk: Extreme and the base-rate note.

Data confidence

High, Medium or Limited. Separate from the score.

Costs & limits

• Scores on today’s top-rated cards and market pages are free. Generated analyses carry the score in their report.

A score is not a probability, a promise or advice. Strong scores on limited data deserve extra caution.