How the Bitcoin On-Chain Dashboard Calculates Its Views
This methodology documents the calculations and display choices in BTC On-chain Lab v1.1.1. The application compares Bitview records; it does not independently index the Bitcoin blockchain.
Select an observation, not an approximation
Requests specify a UTC date, cohort, native lin1000 bucket scheme and raw BTC weight. A requested missing day is not substituted with the closest available day. Cohort freshness is assessed against that cohort’s available dates; current-day data may be incomplete.
Before display the app checks finite fields, date and cohort identity, bucket schema, supply sums and available monetary reconciliation. Invalid snapshots are rejected. Successful validation means the records pass consistency checks, not that the underlying source has been independently audited.
Buckets, uncertainty and profit/loss
Native price ranges are half-open $1,000 buckets. Display steps combine them, while profit/loss classification still uses the native ranges. The native bucket crossing the selected source price remains unclassified. Near-price minimum supply counts wholly enclosed ranges; maximum supply also includes boundary ranges.
BTC interval bounds are rounded outward for display: the minimum down, the maximum up. Exact zero remains zero. A region’s net P&L comes from summed source unrealized_pnl, equal to snapshot price times BTC minus realized cap when the source inputs reconcile. It is not a bucket-midpoint cost estimate.
Two-date changes and percentage denominators
| Display | Calculation / scope |
|---|---|
| Net change | Selected-day BTC minus comparison-day BTC in each range. |
| Added / removed | Positive and negative balance changes, calculated per displayed range. |
| Percentage mode | Both dates divided by selected-day cohort supply. |
| Filtered table total | Only rows actually displayed after range and zero-value filters. |
| Full snapshot export | Retains source rows and metadata even when the chart is filtered. |
Comparison-date percentages can total above 100% when that date’s supply exceeds the selected-day denominator. A wider display step can cancel opposing changes and alter gross additions/removals. Neither amount represents trading volume. Tiny positive quantities may round to zero in the display while raw exports preserve precision.
Network valuation and scenarios
Realized price = realized cap / supply; MVRV = market cap / realized cap; NUPL = 1 − 1 / MVRV. The macro section always uses the network snapshot. Realized cap is taken from the source when valid or derived from complete market cap and unrealized P&L inputs.
The NUPL reference boundaries are 0, 0.25, 0.5 and 0.75, evaluated before rounding. They are descriptive conventions, not an independently backtested investment strategy. Simulation holds supply distribution and realized cap fixed; it excludes behavioral responses and does not predict future spending.
Cohorts and coin age
This source’s STH/LTH split uses less than 150 days versus at least 150 days. Do not substitute another provider’s 155-day rule. Coins can age into another group without a transfer. Address-balance, UTXO-size and other cohorts are source-defined categories, not counts of distinct people.
Coin age measures output dormancy. It does not reveal why coins remained unspent or whether their keys are accessible. Rookie/Veteran categories compare creation price against network capitalized price at creation, weighted by creation-date USD value; they are not another short/long age split.
Reproduce a result
Record both UTC dates, the source cohort ID, selected unit and display step. Open the source snapshot links and export JSON or CSV. JSON includes sourceSnapshots and stepUSD; quantity columns distinguish displayed units from raw BTC. Compare native rows before interpreting aggregated changes.
PNG records the visible chart, while CSV and JSON are better for numerical checks. Data can be delayed, incomplete or revised. Retain source identity when comparing with another service and report unresolved boundary supply alongside classified values.