Read the returned dates, then the fitted bands
The page requests BTCUSD end-of-day prices from July 2010 onward, drops unusable dates and nonpositive prices, and fits log10(price) against log10(days since genesis) by least squares. Check the first and last observation dates and the displayed coefficients. The fit uses only returned rows. It does not fill earlier history.
Nine bands surround the curve. Adjacent boundaries differ by 0.42 log10 units, a display choice with no statistical coverage interpretation. Prices outside all bands are labeled outside the range rather than assigned a buy or sell signal. The latest observation is not a live quote, and extending the curve is not a price forecast.
Read the sample before reading the colors
- Check the first and last observation dates. The latest historical observation is not a live quote.
- Inspect the displayed fitted coefficients and sample size. Revisions or a changed history window can move every band.
- Use the curve-extension control to see extrapolation sensitivity. Extending a mathematical curve is not evidence of future returns.
- Treat halving markers as context only. The fit does not isolate halving effects from other causes.
What the model cannot establish
This is an in-sample descriptive visualization. It contains no out-of-sample test, cash-flow valuation, market-sentiment measurement or profitable trading rule. A line fitted using future observations cannot be treated as a signal that was available earlier in history.
Method reference: NIST on least-squares fitting and extrapolation limits. Bitcoin subsidy timing is documented in Bitcoin Core mainnet parameters.