Posted by LearnBitcoin.com
Nov 11, 2026/19:50 UTC
The introduction of standardized terminology such as PSG and RLA has provided increased clarity when discussing issues surrounding decentralized mining pools, particularly those related to P2Pool. Prior to this, the lack of uniform vocabulary made discussions less precise, as participants described similar phenomena in varying terms. Historical records confirm that the referenced post from November 2015 on BitcoinTalk, authored by windpath, was indeed significant. This individual operated a public, zero-fee P2Pool node at p2pool.org during the relevant years and still maintains control over the domain.
A close examination of every block’s coinbase transaction that paid to P2Pool’s donation script reveals 2,545 such blocks, with over a thousand occurring in 2012 alone. Even today, the pool continues to generate one or so per year. Data shows that, in 2012, a typical P2Pool block distributed rewards to approximately 200 addresses, which grew to about 390 by 2014. The largest distribution occurred on May 19, 2013, when 553 addresses were paid in a single coinbase transaction spanning 19 kB. Only 42 blocks ever rewarded 500 or more addresses. From its inception, the original P2Pool code included a “rich list” rule that retained only the 4,000 largest payouts, labeled as “unlikely to ever be hit”; empirical data supports that this limit was indeed never reached.
Challenges faced by small miners in P2Pool primarily arose from share difficulty adjustments, compounded by the dust rule. When a miner’s expected payout per block dropped below 0.001 BTC, the software automatically raised that miner’s share difficulty until each share exceeded the threshold. Evidence on-chain demonstrates a clear cutoff: for example, in July 2014, out of 15,865 payouts, 1,749 fell between 0.001 and 0.002 BTC, but none dipped below 0.001. Miners unable to meet this threshold simply disappeared from payout records—they are not represented as small outputs because they received nothing and likely ceased participation. Importantly, these outcomes were driven by natural network growth rather than targeted attacks. As articulated in the 2015 post, there are inherent limitations in how many times a fixed block reward (such as 25 BTC) can be divided within a decentralized, trustless system; inevitably, growth alone will squeeze out smaller participants. In effect, the RLA represents the ultimate outcome of any ranked-slot rule, whether due to organic network expansion or deliberate manipulation by an adversary. For those interested in the raw data, a CSV file detailing coinbase payouts is available at https://www.learnbitcoin.com/downloads/p2pool-coinbases.csv.
A proposed potential solution, first outlined in 2015, suggested allowing miners to accumulate shares in a trustless manner, with payouts in BTC issued only once a particular threshold is reached. This design would prevent the creation of dust outputs within the generation transaction. Despite the passage of eleven years, contemporary proposals continue to reflect this basic structure, underscoring the persistent complexity involved in developing a robust solution for the intermediary phases of decentralized payout protocols.
With regard to the question of who might engage in honest mining while simultaneously undermining payout systems through griefing, established pools and large miners have clear incentives. Since they already possess substantial hash power, redirecting a portion of it to a decentralized pool incurs minimal cost—primarily slightly increased latency. Block withholding reduces their income and attracts attention, whereas dividing their hash power among numerous accounts emulates the appearance of increased participation by small miners without additional expense. This tactic allows them to subtly degrade the new pool’s performance, making it consistently less attractive compared to centralized alternatives. As such, payout griefing poses a more effective and deniable threat than outright sabotage.
Key considerations for designing a resilient value-weighted draw system include ensuring that splitting claims does not confer any advantage. If a single large claim occupies only one slot, a miner could still manipulate outcomes by fragmenting their claim into multiple smaller ones, reintroducing the RLA dynamic as a probabilistic challenge. To prevent this, chances of being selected should correlate strictly with total value, irrespective of claim partitioning. Additionally, the source of randomness (the seed) must be carefully chosen. Since coinbase outputs are set before mining, the randomization must derive from prior data, ideally from the preceding Bitcoin block rather than the sharechain, as the latter is more susceptible to manipulation. Furthermore, once the seed is determined, claims must remain immutable to prevent gaming of the system. When these conditions are met, attempts by incumbents to disrupt the protocol merely increase costs for market makers, which is a preferable failure mode compared to indefinite or misdirected payouts.
For those seeking further clarification on decentralized coinbase payout protocols (DCPP), PSG, and RLA, comprehensive glossary pages are
TLDR
We’ll email you summaries of the latest discussions from high signal bitcoin sources, like bitcoin-dev, lightning-dev, and Delving Bitcoin.
We'd love to hear your feedback on this project.
Give Feedback