Statistical transparency
Methodology
How events, independent observations, and subsequent returns are defined.
Daily data and UTC
All source timestamps are converted to UTC before deriving dates. CoinGecko's normalized daily price is the reference point. It is not OHLC data, so intraday highs, lows, and drawdowns are not claimed.
Thresholds and episodes
Threshold comparisons are inclusive. Consecutive or nearby qualifying observations are grouped according to each definition's cooldown, and only the first date triggers the occurrence. A red streak triggers once when it first reaches its requested length. Nearby all-time highs are grouped.
Forward returns
For event date T and horizon H, return is ((price[T+H] / price[T]) − 1) × 100. The exact UTC day is required. We do not silently jump over missing days. An unavailable future date makes that horizon incomplete, and incomplete observations do not enter aggregate statistics.
Statistics
Positive probability is the number of completed returns above zero divided by completed returns. Means are arithmetic. Medians and quartiles use inclusive linear interpolation over sorted outcomes. Best and worst closing moves describe daily reference points, not intraday movement.
Cross-asset studies
Bitcoin determines the trigger date, but each target asset's return starts from its own price on that date. Events before the target asset's data begins are excluded.
Indexing and limitations
Pages need the configured minimum number of completed 7-day observations to be indexable. Historical association is not causation or a prediction. Provider coverage, reference-price construction, survivorship, and changing market structure all limit interpretation.