Methodology — how our models work
Pastatrade is not a black box. Every score is a transparent, rules-based read of market conditions. Below is what each model measures, the data behind it, how it's calculated, and — honestly — its limits.
Principles behind every model
Research, not advice
Every score describes market conditions. It never says “buy” or “sell.”
Probabilities, not predictions
We map risk and structure — not where price goes next.
Plain language
Every score becomes a clear, everyday sentence anyone can act on.
Graceful degradation
If a data source is down, the model reweights over what’s available and lowers confidence.
Relative strength matters
Altcoins are judged against Bitcoin (the BTC pair), not just USD.
Confidence is shown
We tell you when a read is weak, incomplete, or based on thin data.
The models
BTC Risk & DCA
Scale: 0–1- What it measures
- How cheap or overheated Bitcoin is on a blended risk model.
- Inputs / data
- Drawdown from all-time high, distance from long-term moving averages, RSI, and on-chain context.
- How it's scored
- Each factor is normalised to 0–1 and blended into one score. Low = accumulation-friendly (DCA zone), high = distribution risk.
- Labels
- Good DCA zone (<0.35) · Neutral (0.35–0.55) · Caution (0.55–0.75) · Distribution risk (>0.75).
- Limitations
- A risk model, not a price prediction. It describes conditions, not timing.
Data sources & refresh
Models are built from public market data — CoinGecko, blockchain.com, DefiLlama, Fear & Greed (alternative.me), Google Trends, Wikipedia, YouTube, Bitget, GeckoTerminal and GoPlus. Data is refreshed on a scheduled daily sync; when a source is temporarily unavailable the affected model shows “unavailable” or a reduced-confidence read rather than a stale number. Reports summarise all of the above in plain language (English & Swahili).