On-chain data has a seductive property: it is complete and verifiable, which makes it feel objective. The data is objective. The metrics built on it embed assumptions, and most of those assumptions are wrong in the same few ways.<\/p>\n
The most common error and the foundation of several widely-cited metrics.<\/p>\n
One person can control thousands of addresses. One address can hold assets for millions of people. Exchanges hold customer balances in a small number of addresses; a single individual using privacy-conscious practices generates a new address per transaction.<\/p>\n
\u00abActive addresses\u00bb therefore does not measure users. It measures address activity, which correlates with usage loosely and breaks down exactly when behaviour changes \u2014 which is when you most want the metric.<\/p>\n
Clustering heuristics attempt to group addresses by common ownership, and they are useful. They are also probabilistic, they degrade as wallet software improves, and any metric built on them inherits their error rate silently.<\/p>\n
Transfer volume counts value moved, including value that was not going anywhere meaningful.<\/p>\n
Exchanges rebalancing between their own wallets. Bridges moving assets between contracts. Automated strategies cycling positions many times a day. A single economic transaction generating several on-chain transfers as it routes through contracts.<\/p>\n
Adjusted volume metrics attempt to filter this, and the adjustments are themselves heuristics with their own assumptions. The useful practice is to ask what an adjustment removed and whether you agree with the removal \u2014 the methodology is more informative than the number.<\/p>\n
\u00abExchange inflows\u00bb and \u00aboutflows\u00bb are among the most cited on-chain indicators, and they depend entirely on knowing which addresses belong to which exchange.<\/p>\n
Those labels are inferred \u2014 from clustering, from disclosed addresses, from observed patterns. They are incomplete, they go stale when an exchange rotates infrastructure, and different data providers produce materially different figures for the same exchange on the same day.<\/p>\n
The metric is not useless. It is a measurement with an error bar that is rarely published, and a large apparent flow can be an exchange reorganising its own wallets.<\/p>\n
TVL is a quantity of assets multiplied by their price. Both move, and price usually moves more.<\/p>\n
A protocol whose TVL fell by half may have lost half its deposits, or may have held every deposit while the assets fell in value. These are entirely different events and the headline number does not distinguish them.<\/p>\n
Double-counting compounds it: an asset deposited in one protocol, tokenised as a receipt, and deposited in a second counts in both. Aggregate figures across an ecosystem can include the same underlying capital several times.<\/p>\n
The more informative version is denominated in units of the underlying asset rather than in currency, which separates deposit behaviour from price movement.<\/p>\n
\u00abTop ten addresses hold X percent\u00bb is frequently cited as evidence of concentration. Those addresses are often exchange wallets, bridge contracts, staking contracts or treasury multisigs \u2014 none of which is a holder in the sense implied.<\/p>\n
Conversely, genuine concentration can be hidden across many addresses controlled by one party. The metric can overstate and understate concentration simultaneously, in the same dataset.<\/p>\n
Ask what the metric literally counts, not what it is named. Ask what assumption converts that count into the claim being made. Ask what would make the metric wrong, and whether that condition is present. Prefer denominating in underlying units over currency where you are trying to observe behaviour. And compare providers \u2014 where two reputable sources disagree materially on the same metric, the disagreement is the finding.<\/p>\n
None of this is investment advice or a claim about any current market condition. It is a description of where the common metrics break, which is knowledge worth having before relying on one.<\/p>\n","protected":false},"excerpt":{"rendered":"
Why on-chain metrics mislead: addresses are not users, volume is not economic activity, and the clustering assumptions behind most dashboards.<\/p>\n","protected":false},"author":10,"featured_media":145,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[13],"tags":[89,76,90,88],"class_list":["post-116","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-market-trends","tag-analytics","tag-market-structure","tag-methodology","tag-on-chain-data"],"_links":{"self":[{"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/posts\/116","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/users\/10"}],"replies":[{"embeddable":true,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/comments?post=116"}],"version-history":[{"count":1,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/posts\/116\/revisions"}],"predecessor-version":[{"id":122,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/posts\/116\/revisions\/122"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/media\/145"}],"wp:attachment":[{"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/media?parent=116"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/categories?post=116"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tokenpr.net\/es\/wp-json\/wp\/v2\/tags?post=116"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}