Methodology
Sentiment Classification
Every piece of content is analyzed by our AI pipeline using large language models (Claude) to extract directional sentiment and relevant assets.
The model evaluates context, considers hedging language, and distinguishes between strong conviction calls and nuanced observations. Each classification includes asset extraction (BTC, ETH, SOL, etc.) for granular analysis.
Consensus Index
The Consensus Index uses a one-analyst-one-vote system. Each analyst's most recent post (within 30 days) counts as their vote, weighted by their historical balance score—analysts who perform well in both bull AND bear markets carry more influence.
We apply 30-day and 90-day moving averages to smooth out daily noise. Chart colors use 90-day rolling Z-score to highlight unusually bullish or bearish periods relative to recent history:
- Green = Above-average bullish sentiment
- Red = Above-average bearish sentiment
Accuracy Score
We evaluate analysts using a phase-based scoring system. Market history is divided into predefined bull and bear phases, and each analyst's calls are evaluated against these ground-truth periods.
- At least 50 days of sentiment data
- Participation in at least 2 bull AND 2 bear phases
- Active posting history (not inactive for 6+ months)
- At least 1 year of history from first to last data
- Both long and short phase participation for Long/Short-term categories
Note: We only analyze the most recent 10,000-20,000 tweets per analyst. For very active accounts with many tweets, this may not cover their full history, potentially affecting accuracy calculations.
Market Phases
Analyst accuracy is evaluated against predefined market phases. These phases are automatically detected from BTC price data using a swing-based algorithm, ensuring objective and consistent ground truth.
Our algorithm analyzes weekly BTC price data to identify significant market turning points:
- Detect peaks and bottoms using a 30% swing threshold
- Generate phases between consecutive extremes
- Merge consecutive same-direction phases
- Classify phases by volatility (under 40% = allows neutral)
Each phase is classified based on price direction and volatility:
View Phase Details
| Start | End | Days | Correct Sentiment |
|---|---|---|---|
| 2020-09-07 | 2021-04-12 | 217 | BULLISH |
| 2021-04-12 | 2021-07-19 | 98 | BEARISH |
| 2021-07-19 | 2021-11-08 | 112 | BULLISH |
| 2021-11-08 | 2022-11-21 | 378 | BEARISH |
| 2022-11-21 | 2024-03-11 | 476 | BULLISH |
| 2024-03-11 | 2024-08-05 | 147 | BEARISH or NEUTRAL |
| 2024-08-05 | 2024-09-02 | 28 | BULLISH or NEUTRAL |
| 2024-09-02 | 2024-09-09 | 7 | BEARISH or NEUTRAL |
| 2024-09-09 | 2024-12-17 | 99 | BULLISH |
| 2024-12-17 | Present | — | TBD |
Analyst Bias
Current Bias measures an analyst's recent directional leaning based on their last 5 posts within a 30-day window. This provides a real-time snapshot of their current market stance.
The bias score ranges from 0 (perma-bear) to 100 (perma-bull), with 50 being neutral. Nuanced calls count as half weight. Analysts with no directional posts in 30 days show 'Inactive'.