Skip to content

Merged Metrics Data Guide

Generated from scripts/generate_merged_metrics_taxonomy.py against merged_metrics_2026-03-15_04-29-57.parquet.

Use this page first if you want to understand what data you actually have access to through merged_metrics*.parquet.

Reading order:

  1. Merged Metrics Data Guide
  2. Merged Metrics Parquet Schema
  3. Merged Metrics Taxonomy

What This Dataset Is

The canonical BRK dataset is a long-format daily fact table with exactly three columns:

  • day_utc: UTC calendar day
  • metric: metric key
  • value: numeric value for that metric on that day

Each row is one daily numeric observation for one Bitcoin analytics metric. The dataset gives you access to thousands of derived BTC time series, not a wide table of fixed columns and not raw transaction-level blockchain records.

Current snapshot scale:

  • 236,259,020 rows
  • 6,274 daily observations
  • 41,407 metric keys
  • 2009-01-03 to 2026-03-13 coverage in the current snapshot

What Data You Can Access

The metric namespace covers these major user-facing domains:

Domain What it covers Representative metrics Typical use Coverage notes
Address balance cohorts Address cohorts partitioned by balance bucket. addrs_above_100btc_under_1k_btc__30d_change, addrs_above_100btc_under_1k_btc__30d_change_btc, addrs_above_100btc_under_1k_btc__30d_change_usd Whale, retail, and cohort-balance distribution analysis. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Benchmarks, path metrics, and technical indicators Windowed return paths, DCA/lump-sum ladders, and indicator-style metrics. 10y_cagr, 10y_dca_average_price, 10y_dca_average_price_sats Benchmarking, path-dependent outcomes, and technical overlays. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Blocks, transactions, and network activity Blocks, transactions, throughput, activity, address counts, and network utilization. _30d_change, _30d_change_btc, _30d_change_usd On-chain activity monitoring and network-usage context. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Holder cohorts Short-term and long-term holder slices and holder-behavior metrics. lth__30d_change, lth__30d_change_btc, lth__30d_change_usd Compare STH and LTH behavior across market regimes. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Market and valuation Price, valuation bands, market value, realized value, and valuation ratios. active_cap, active_price, active_price_ratio Price context, valuation regimes, and long-horizon market state modeling. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Mining pools and miner economics Mining-pool shares plus miner economics such as hash-price and fee flows. aaopool_1m_blocks_mined, aaopool_1m_dominance, aaopool_1w_blocks_mined Miner revenue, pool share, and mining-economics monitoring. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Profitability and SOPR Realized and unrealized profit or loss, SOPR-style metrics, and spending pressure. adjusted_sopr, adjusted_sopr_30d_ema, adjusted_sopr_7d_ema Profit-taking, capitulation, and spending-behavior analysis. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Script and output types Output/script-type cohorts including p2* families, unknown, empty, and OP_RETURN. empty_addr_count, empty_addr_count_30d_change, empty_outputs__30d_change Track script adoption, output composition, and script-specific cohorts. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Supply, issuance, and scarcity Circulating supply, issuance, subsidy, inflation, and scarcity metrics. active_supply, active_supply_btc, active_supply_usd Supply-side modeling and issuance or dilution analysis. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
UTXO age cohorts Age-bucketed UTXO metrics by holding period cohort. utxos_10y_to_12y_old__30d_change, utxos_10y_to_12y_old__30d_change_btc, utxos_10y_to_12y_old__30d_change_usd Analyze age-distributed supply, spentness, and conviction. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.
Vintage and halving cohorts Year-vintage and halving-epoch cohort metrics. epoch_0__30d_change, epoch_0__30d_change_btc, epoch_0__30d_change_usd Cycle-aware cohort comparisons and halving-era analysis. 2009-01-03 to 2026-03-13 across at least some metrics; per-metric coverage still varies.

Major Data Domains

High-level examples of what a new user can query from the metric namespace:

  • Market and valuation: price_*, market_*, realized_*, mvrv, investor_*, cost_*
  • Profitability and SOPR: *_sopr*, *_profit*, *_loss*, capitulation_*, pain_*
  • Supply and scarcity: supply_*, circulating_*, subsidy_*, inflation_*
  • Holder cohorts: sth_*, lth_*
  • UTXO age cohorts: utxos_<age_bucket>_*
  • Address balance cohorts: addrs_<balance_bucket>_*
  • Vintage and halving cohorts: year_<yyyy>_*, epoch_<n>_*
  • Mining pools and miner economics: <pool>_blocks_mined, <pool>_dominance, hash_price_*, coinbase_*, fee_*
  • Script and output types: p2*_*, unknown_*, empty_*, opreturn_*, segwit_*, taproot_*
  • Blocks, transactions, and network activity: block_*, tx_*, hash_rate, difficulty*, sent*, received*
  • Benchmarks and path metrics: 1m_*, 1y_*, 10y_*, dca_*, rsi_*, macd_*

What This Dataset Does Not Contain

  • It does not expose raw transaction rows, raw block rows, or raw address ledgers.
  • It does not provide intraday timestamps; observations are daily.
  • It does not make every metric a dedicated parquet column; access happens through metric keys in long format.
  • StackSats runtime does not consume all 41,407 metrics directly. Runtime uses a smaller derived BRK-wide projection.

Coverage Caveats

  • Coverage is metric-specific. Some metrics begin much later than 2009-01-03.
  • Newer transforms and ratios can have shorter history because they depend on warmup windows or derived inputs.
  • Use data/brk_merged_metrics_catalog.json to inspect coverage_rows, first_day, and last_day for each metric.

Metrics Used By StackSats Runtime

These runtime-critical metrics are the minimum projection used by built-in strategy audit tooling:

Metric Coverage rows First day Last day
market_cap 6,274 2009-01-03 2026-03-13
supply_btc 6,274 2009-01-03 2026-03-13
mvrv 6,273 2009-01-09 2026-03-13
adjusted_sopr 5,689 2010-08-16 2026-03-13
adjusted_sopr_7d_ema 5,689 2010-08-16 2026-03-13
realized_cap_growth_rate 5,324 2011-08-16 2026-03-13
market_cap_growth_rate 5,324 2011-08-16 2026-03-13

The runtime projection renames day_utc to date and derives price_usd from market_cap / supply_btc. See Merged Metrics Parquet Schema and BRK Data Source.

How To Search The Catalog And Taxonomy

  • Use data/brk_merged_metrics_catalog.json when you want per-metric access metadata and coverage.
  • Use Merged Metrics Taxonomy when you want family-level naming patterns.
  • Search by access_category to find domains, by family to find namespaces, and by display_label or metric when you already know the concept or key.
  • If a metric name looks inconsistent, the catalog notes field explains metadata-only normalization such as collapsed double underscores or family aliases.