BITCOINREVERSAL RESEARCH / Original research
Bitcoin Opportunity Score
A daily 0–10 ranking of Bitcoin’s historical risk/reward structure.
Bitcoin price and Opportunity Score chart
Hover or tap to explore. Higher scores rank more favorably in the frozen model. Historical scores were reconstructed later; they are not live signals published at those dates.
How to interpret the Bitcoin Opportunity Score
At 4.94 / 10 on Sep 19, 2026, the model ranks Bitcoin’s structure as neutral. Higher values rank more favorably against the frozen training distribution; 7 / 10 does not mean a 70% chance of gains.
These four groups explain the latest reading. Positive contributions support the raw model output; negative contributions weigh on it. They are raw model units and do not add up to the final 0–10 score.
BTC is 0.7% above its 365-day average and 34.9% below its known ATH.
The 365-day average changed -12.2% over 90 days; 348 days have passed since the known ATH.
Trailing 30-day realized log-return volatility is 43.3% annualized.
Recovery gap is 43.6 log-points, comparing trend position with distance from ATH.
The ranking is continuous. The highlighted zones are reference ranges, not buy or sell instructions.
Historical testing of Bitcoin Opportunity Score
Recovery-6 uses coefficients fitted to earlier data. The later-period diagnostic uses the first Monday of each month from January 2023 to September 2025: 33 observations with complete 365-day outcomes. These are retrospective tests, not a prospective track record.
Median 365-day return in the later-data diagnostic.
Median 365-day return in the same diagnostic.
Reproduced from the published model and stored daily history on September 8, 2026. Inspect all 33 observations and the calculation method.
Retrospective examples
Model designed later; examples include a notable miss.
BTC ~$16.7k · next 365d +169%
Attractive reconstructionBTC ~$44.2k · next 365d +112%
Attractive reconstructionBTC ~$5.0k · next 365d +1,033%
Important failure caseHow the Recovery-6 model is calculated
Six price-derived inputs feed a fixed statistical model. Its output is ranked against the training distribution. New completed daily closes update the reading; the coefficients stay fixed.
Data sources and daily timing
The chart uses completed UTC calendar days. From March 14, 2026, daily closes are the last CoinGecko BTC/USD price sample recorded by BitcoinReversal that day. Earlier rows are imported history; their original upstream provider and closing convention have not been verified.
CoinGecko price documentationFrozen model specificationSource and calculation metadata
Calculation details, version history & publication log
Reproduce a reading
Download the complete daily history and the model specification. For each date, calculate the six features above, subtract each published scaler mean, divide by its scaler scale and multiply by its coefficient. Add the intercept. Linearly interpolate this raw prediction against the 101 frozen quantile thresholds and divide the percentile rank by ten. The first score needs 455 daily observations; the published score series starts April 2, 2012.
Volatility uses 30 daily log returns and population standard deviation. Time since ATH is log(1 + elapsed days) / log(1461); a close matching the known high resets the counter. Recovery gap is the log price/SMA365 distance minus log price/ATH distance. Four displayed contributions group six terms before the final quantile conversion.
Closing-time and revision limits
A recorded daily close is the final collected sample, not a guaranteed exchange settlement price or an exact 23:59:59 trade. Collector outages and historical corrections can affect readings. The current incomplete UTC day is excluded. Weekly and monthly controls only change which completed daily observations are displayed; they do not refit the model.
Model and presentation versions
Recovery-6 v1: the published coefficients and quantile thresholds are fixed. “Pre-2023 training” describes the training-outcome cutoff, not the date this model was created or first published. September 8, 2026: presentation, source disclosures and export improvements; no changes to the score formula, coefficients or training thresholds.
Forward publication record
The publication log begins with this release. Each entry records the observation date, first recorded publication time, model hash and exact values. Entries are not retrospectively backfilled. This is a publisher-maintained record, not independent timestamp certification or proof that older reconstructed scores existed at the time.
Input history JSONForward publication logReport a data or methodology issue
Model limitations and failure cases
A credible score is useful only if its failure modes stay visible.
Not a probability
7.0 / 10 does not mean a 70% chance of gains. It is a model ranking.
Not prospective history
The historical score line is reconstructed with a model designed later.
Overlapping outcomes
Monthly 365-day tests share much of the same future path.
Regimes can change
ETFs, derivatives, regulation and volatility can alter historical relationships.
Shock recoveries can be missed
March 2020 is a visible example of a defensive score before an extraordinary rebound.
Price data only
Recovery-6 excludes macro, flows, on-chain metrics, fees, taxes and personal constraints.
A reading you can trace.
Use the model name, version and observation date when citing a score. Link to this canonical page so readers can check the data and limitations.
Bitcoin Opportunity Score questions
What is the Bitcoin Opportunity Score?
BitcoinReversal’s experimental 0–10 score ranks current Bitcoin market structure using six price-derived Recovery-6 inputs and a frozen model trained on outcomes ending before 2023.
Is a high score a buy signal?
No. A high value means the model ranks the structure more favorably relative to its frozen historical training distribution.
Does the historical score repaint?
The chart is a retrospective reconstruction. Completed values change only if the underlying price history or model specification changes.
Why not include the Bitcoin halving cycle?
The published Recovery-6 specification uses six price-derived features and does not include halving-calendar inputs. Its calculation is independent of scheduled halving dates.
How often is it updated?
The reading updates when a new completed UTC day becomes available. The current incomplete day is excluded; the date above the chart tells you exactly which observation is shown. A delayed-data label appears if the previous day is missing.
Is this part of the paid BitcoinReversal signal?
No. This public research model is separate from the proprietary BitcoinReversal indicator framework and does not expose or replace its signal logic.
Compare independent Bitcoin cycle lenses.
Opportunity Score is designed to complement—not replace—the other public Bitcoin models.