Mining Pool Centralization means pooled mining or decisions about it become concentrated among a small number of pools. It has several forms: shared transaction selection, a shared treasury and common ownership of machines are different dependencies. A pool’s share of blocks on a chart is the start of an investigation, not its conclusion.
A Mining Pool combines miners’ work and distributes rewards among them. A miner can receive a steadier income without waiting to find a whole block alone. Their ASIC chips, specialized mining hardware, can still belong to them. A large pool share does not imply equivalent ownership of machines. Eva therefore distinguishes the equipment owner, whoever can change its settings, and the operator preparing work and managing payouts. Sometimes these roles overlap; sometimes different businesses perform them. Concentration needs describing for each role. [1] [4]
A block does not list every machine that worked on it. The mempool project identifies pools through coinbase_tags and payout_addresses, known tags and reward addresses. It looks for them in coinbase, the special initial transaction creating the block reward, unrelated to the exchange of that name. The chart’s label is therefore an attribution based on known clues. It does not establish hardware ownership. If clues are missing, leave the block unknown. Assigning it to the largest pool would insert our own assumption into the data. [2]
Consider our own example: out of 100 blocks, we attribute 40 to pool A, 20 to pool B, 20 to other known pools and leave 20 unknown. A+B account for 60% of all blocks, but 75% of identified blocks only. No miner switched pools; we changed which blocks count toward the total. This is not the network’s current state. For a real chart, Eva checks the period, sample size and treatment of unknown blocks. Discoveries are random, so a short window is not an exact inventory of physical computing power either. [1] [2]
Eva revisits the same period and sees a different breakdown. That need not indicate a blockchain error. The mempool project’s introductory documentation, its README file, warns that incomplete renaming can split one pool into two entries. Updating the mapping in pools-v2.json may also lead to reassignment of previously unknown or affected blocks. A chart’s change may therefore be a data correction, not miners moving. A fair comparison records both block dates and the version of attribution and name-merging rules. [2]
A block template proposes which transactions a block should contain. If one operator prepares it for many machines, an important decision becomes concentrated there. The communication protocol Stratum V2 can move that choice closer to the miner, but its Mining Protocol alone can still distribute pool-selected work. Independent selection requires deployed Job Declaration: JDC is the miner-side software component, JDS its counterpart for the pool, and the Template Provider supplies proposed blocks. The Stratum V2 label alone therefore does not prove decentralized template creation. Eva asks what the operation actually runs. [3] [4]
Job Declaration offers different modes. In Coinbase-only, the pool need not know the miner’s transaction set; in Full-Template, JDS can obtain its data. That privacy concerns work preparation: once a block is found and published, its transactions are public. Here a token is an identifier allocated for a mining job, not a tradable coin. If cooperation fails, for example through a rejected declaration or an unanswered token request, the specification requires JDC to switch to another Pool+JDS, or ultimately solo mining, working independently. Switching requires a working alternative. It does not prove all miners can leave instantly. And changing who selects transactions does not by itself divide money management among independent operators. [4]
Concentration brings real risks: coordinated action can increase the ability to omit particular transactions or reorganize recent block history. Adding shares on a chart does not itself prove agreement between pools, however. Section 11 of the Bitcoin whitepaper also explains a limit to this power: even building an alternative chain faster does not authorize creating coins outside the rules or spending someone else’s money without permission. Nodes, the computers checking Bitcoin’s rules, reject invalid transactions. Coordination deserves scrutiny, but it is not unlimited control of the network. [5]
Finally, Eva wants to know where the machines stand. Neither a pool name nor a coinbase tag reveals that. Cambridge’s 2025 report used data from 49 mining firms representing 48% of hashrate, the network’s computing power, at the time of collection. It also warned that stronger participation by American firms may have overstated the US share. This is a useful operational sample, not a complete map of today’s network. Understanding the picture requires separate evidence about work coordination, payouts, ownership and equipment location. One pie chart cannot answer all those questions. [6]
For the clearest picture, read this entry together with Mining Pool, Hashrate, Job Negotiation, Proof of Work, Bitcoin, Stratum V2. The reverse links also lead from The 7th Property.