BTC: Observed Hash Rate by Mining Label

Bitcoin Observed Hash Rate by Mining Label: daily data through 7 Nov 2026.

Difficulty-derived statistical work per second split by mining-label category, in EH/s.

Why use it. Compare observed work estimates under the same retrospective attribution rule. How to read it. This is not measured hardware capacity or verified operator control. Short periods are noisy, and Decimal arithmetic cannot recover precision lost in the retained native difficulty estimate. An empty sealed interval can contribute zero estimated work. The chart, the series API and the CSV/JSON exports select the 12 largest named labels by block count in the selected range and group the remaining named labels as Other labelled pools, with the same exact values. Unknown and Ambiguous stay separate. The paged raw-history download keeps every contributing label. The pinned label list is applied retrospectively. Tags can be spoofed, and different labels can belong to the same operator. Read label coverage alongside these observations. Read actual source coverage and clocks. Open periods remain partial; MTP closure and six later canonical blocks govern finality. The calculation describes unsmoothed observations; chart smoothing is a separate transform.

Calculation. Sum each category's retained relative block difficulty multiplied by 2^32; divide by the full 3,600-second hour or 86,400-second day and by 10^18 for EH/s. Do not substitute physical hash attempts or exact Bitcoin Core chainwork. Match literal UTF-8 coinbase-tag bytes with ASCII-only case folding, together with the exact canonical address of coinbase output 0. One candidate gives a named label, none gives Unknown and conflicting candidates give Ambiguous. Assign complete canonical blocks to UTC hours or days using their header timestamps, retaining timestamp regressions. Use locally validated canonical coinbase facts and the pinned MIT-licensed mining-pool definitions; credits and the dataset revision remain in method details.