On-chain Analysis

Bitcoin Developer Guide describes inputs referencing earlier outputs and new outputs with an amount and spending conditions. Our analysis separates these verifiable data from interpretations of purpose: the record alone does not say whether this was a purchase of goods, a transfer between one’s own wallets or another operation. A transaction identifier is not a participant’s civil identity. [1]

Bitcoin Developer Guide explains change as an ordinary transaction output. Our measurement therefore does not label the entire output sum as a payment to another owner. Change and transfers between one’s own addresses can increase gross volume without the same amount changing economic owner. Recognising such transfers requires additional evidence or explicit assumptions. [1]

Glassnode groups addresses into estimated entities using heuristics and describes these data as statistical and mutable. Our interpretation therefore does not turn active-address counts into people counts. One actor can use many addresses, and a custodian can hold funds for multiple clients. Estimated common control is neither identity verification nor an exact census of Bitcoin users. [3]

Lightning Network allows allocations of funds in payment channels to be updated without recording every individual payment on the blockchain. Our on-chain measurement therefore does not cover all payments. A blockchain transfer must likewise be distinguished from a service’s internal accounting. The absence of a separate on-chain transaction is not evidence that no economic activity took place. [6]

Glassnode warns of incomplete address coverage, internal movements and subsequent changes to exchange labels. Its on-chain exchange metrics are not spot or derivatives trading volume. Our interpretation does not treat an exchange inflow as evidence of a completed sale. The observed balance also lacks complete liabilities and all assets, so it does not by itself establish the service’s solvency. [4]

Glassnode distinguishes the time assigned to a measured interval from the time a metric is published. Confirmations, processing and reorganisations can change data availability or values. Our records therefore note the end of the observed period, the data acquisition time and the chain state. A chart timestamp alone does not prove that the analyst then knew the final value displayed. [5]

Glassnode describes revisions following changes to clusters, labels or calculations; its Point-in-Time data preserve the value at first publication. Our historical simulation uses the versions available at the time when evaluating decisions made then. Today’s recalculated series may contain knowledge obtained later and introduce Look-ahead Bias. A frozen version, however, does not guarantee a correct ownership estimate. [5]

Bitcoin Developer Guide distinguishes transaction inputs, outputs and unspent-output state. Our methodology records the block range, units, filters, calculation version and label origins alongside the data. It adds price data as a separate external input. On-chain Analysis can describe observed behaviour with stated limitations; by itself it does not provide a certain forecast of Bitcoin’s future price. [2]