Methodology
How we build our probability scenarios
Every analysis ends with three numbers — a Bullish, Neutral and Bearish probability that add up to 100%. Here is exactly how we get there, and what they do (and do not) mean.
Example outlook
1. We start from credible sources
Every analysis begins with primary reporting from established crypto outlets. We do not copy articles — we read the original, link back to it, and write our own analysis. If a claim cannot be sourced, it does not shape the scenario.
2. We break the story down
For each story we identify what kind of event it is (an unlock, an ETF flow, a regulation, a hack…), which coins it touches, and over what time horizon it is likely to matter. We weigh historical precedent, current market conditions, and who actually benefits or loses.
3. We assign three probabilities
We describe a realistic Bullish case, a realistic Bearish case, and the Neutral / most-likely case — then attach a probability to each. The three always sum to 100%. A higher number means we think that outcome is more likely; it is never a promise.
4. AI writes, with risk-based editorial review
Our AI analyst writes the analysis and the probability split, and every story passes automated checks: source corroboration via web search, a reliability rating, and an originality test. Routine market analysis that passes all checks is published automatically and clearly labeled as AI-generated. Anything sensitive — regulation, hacks, fraud, lawsuits, or any story naming a party negatively — is always held for a human editor, who also reviews published coverage on an ongoing basis and can correct or withdraw any piece.
5. We hold ourselves accountable
Every 30-day call is automatically graded against what the market actually did, and published — hits and misses alike — on our public Track record page. Each article also lists the concrete conditions that would change our view. We would rather be transparent and occasionally wrong than confident and misleading. No hype, no 'to the moon' headlines.
Asset sentiment & the Scenario Movement Tracker
Beyond any single analysis, we track how bullish our outlook is on each major asset — and how it shifts over time. This is what powers the Scenario Movement Tracker on the home page, and it follows the same transparent, evidence-first principles.
What the percentage means
Each asset's headline number is the average Bullish-scenario probability across our published analyses that are primarily about that asset, from the most recent 90 days. It measures conviction, not price: a 60% reading means our recent analyses of that asset, on average, assign a 60% probability to the bullish case — it does not predict where the price will trade. The figure is a probability, never a target.
Which analyses are included
The score is calculated using published analyses from the most recent 90 days for which the asset is the primary subject — the asset the analysis is mainly about. Analyses that merely list it as a secondary affected asset are not counted, and the market-wide Daily and Weekly Outlooks are excluded too. This measures how much research we have done ON the asset, not how often it appears in research. Older analyses automatically expire, so the indicator reflects the platform's current outlook rather than historical opinions; assets we have not analysed recently do not appear.
How the analyses are combined
We use a simple, equal-weighted average: every qualifying analysis counts the same. We deliberately do not weight by recency, confidence or popularity, which keeps the figure transparent and hard to game. The Bullish, Neutral and Bearish probabilities are averaged independently and then re-balanced so they always sum to 100%.
How changes are measured
Once a day, at a fixed time (13:00 UTC), we record each asset's reading. The Today figure is live and recalculated whenever a new analysis is published, while Yesterday, the 24-hour and 7-day changes, the trend and the history sparkline are all measured against these daily snapshots. Movement only appears after at least two days of snapshots exist, and new 30-, 90- and 180-day sentiment highs and lows are flagged automatically once enough history has accumulated.
Direction and strength
Alongside the change we show a Status that follows the dominant scenario — whichever of Bullish, Neutral or Bearish currently has the highest probability. When that leading scenario commands a clear majority (60% or more) we mark it Very (Very Bullish / Very Bearish). So a coin whose Neutral scenario leads reads Neutral, not Bearish. The trend tells you which way sentiment is moving; the status tells you which scenario currently leads, and how decisively.
Research base
Every score also shows its research base — the number of analyses from the last 90 days that are primarily about that asset. A reading built on 24 dedicated analyses rests on far more evidence than one built on 3, so we display the count next to each asset. It answers “how much research have we done on this asset?”, not “how often has it been mentioned?”.
Confidence
From that count we derive a simple Confidence label: Low (1–3 analyses), Moderate (4–9), High (10–19) or Very High (20+). Confidence is a coverage indicator — it shows how broad our analytical coverage of an asset is. Higher confidence means a larger body of analysis behind the score; it does NOT mean the prediction is more likely to be correct.
Updated
Each asset also shows when it last received a qualifying analysis — for example “3 hours ago” or “12 days ago” — with a freshness label (Fresh, Recent, Aging or Stale). This tells you how current the score is: a reading drawn from analyses published this week carries more weight than one whose most recent analysis is already weeks old.