{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreig6mmaunkeo4xyjgyrxbwfokohjh3thh4lyrtixwetceajyxiggtu",
    "uri": "at://did:plc:ep3fxt5imrl3vy62u2j2i4vl/app.bsky.feed.post/3mg6f3htshgj2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreigdegaskhagg4noxeor4zneqnou724vago5y7ell7tipeaxbwpp2e"
    },
    "mimeType": "image/png",
    "size": 2333712
  },
  "description": "In the last piece, we covered the threat model. The paper lists the main ways a PoBU system can fail or be weakened in practice: issuer concentration, compromise/coercion, availability, and privacy/linkability.\n\nNow comes the part many readers care about most:\n\nWhat can we actually measure on a live network?\n\nBecause PoBU verification of uniqueness and liveness is enforced off-chain, the paper focuses on what can be checked publicly: chain-derived measurements that anyone can reproduce.\n\n\nWhat t",
  "path": "/pobu-measured-on-a-running-chain/",
  "publishedAt": "2026-03-03T18:20:38.000Z",
  "site": "https://blog.humanode.io",
  "textContent": "In the last piece, we covered the threat model. The paper lists the main ways a PoBU system can fail or be weakened in practice: issuer concentration, compromise/coercion, availability, and privacy/linkability.\n\nNow comes the part many readers care about most:\n\nWhat can we actually measure on a live network?\n\nBecause PoBU verification of uniqueness and liveness is enforced off-chain, the paper focuses on what can be checked publicly: **chain-derived measurements** that anyone can reproduce.\n\n## **What the paper measures**\n\nThe paper uses **two necessary, but not sufficient, signals** to evaluate whether participation is broad and whether control is concentrating.\n\n### **1) Validator-set breadth and dynamics**\n\nThis is about the active validator set over time.\n\nThe paper measures things like:\n\n  * how many active validators exist per session\n  * how many enter and exit\n  * how much overlap there is between sessions\n  * and a churn rate\n\n\n\nThe point is simple: if participation stays broad at the key level, the validator set does not collapse into a tiny, fixed group. The paper treats validator-set collapse and near-zero churn as warning signs.\n\n### **2) Block production concentration**\n\nEven if a chain has many validators, control can still concentrate if a small subset produces most blocks. So the paper also measures block production distribution.\n\nIt uses:\n\n  * top-k author share (top 1 / 5 / 10)\n  * HHI\n  * Gini\n\n\n\nThese metrics describe how spread out block production is across authors.\n\n## **Why these measurements matter**\n\nPoBU’s core claim is about eligibility bounded by unique humans, but the paper stays careful: these measurements are computed over **on-chain keys** , not over verified unique humans.\n\nSo the paper treats these measurements as necessary signals at the chain boundary.\n\nIt also says what would look like failure in practice:\n\n  * validator-set collapse or near-zero churn\n  * high author concentration, where a small number of keys produce most blocks.\n\n\n\n## **How the paper gets the data**\n\nThe paper’s approach is simple and auditable.\n\nIt uses public chain data accessed via WebSocket RPC and extracts data via direct Substrate RPC.\n\nThe extraction process includes:\n\n  * converting timestamps to block ranges\n  * extracting block authors across block ranges\n  * extracting session validator sets\n  * computing concentration metrics from those datasets.\n\n\n\nIt also records endpoints, commands, datasets, and checksums so the measurements can be regenerated.\n\n## **Multiple time windows, not one snapshot**\n\nThe paper evaluates multiple windows to reduce cherry-picking and to separate recent behavior from longer-horizon regimes.\n\nThese are:\n\n  * W1: last 90 days\n  * W2: last 365 days\n  * W3: long-horizon snapshot.\n\n\n\nThe idea is to see how these participation and concentration signals look across different time ranges.\n\n## **A few concrete results the paper reports**\n\nHere are a few of the headline numbers the paper gives.\n\n### **Recent 90-day window (W1)**\n\n  * top-10 block authors produced **4.5% of sampled blocks**\n  * **405** unique block authors appeared in the sample\n  * average validators per session: **258.4** (min **169** , max **367**)\n\n\n\n### **Mid-term window (W2)**\n\n  * average validators per session: **1044.6** (min **169** , max **1767**)\n\n\n\nThese are key-level observations, but they are the kind of public evidence the paper uses to evaluate whether visible consensus control is concentrated or broadly distributed.\n\n## **Limits the paper is clear about**\n\nThe paper does not claim these measurements directly prove unique-human counts.\n\nIt states the limits clearly:\n\n  * the metrics are computed over **on-chain keys** , not verified unique humans\n  * block author statistics are computed from sampled blocks, and sampling can introduce bias, so step sizes are disclosed and treated as a limitation.\n\n\n\nSo the paper treats this as reproducible evidence at the chain boundary, not as a full measurement of identity-layer performance.\n\n## **What comes next**\n\nThe paper also notes that the evaluation would be strengthened by more publishable aggregates from the identity layer, but it keeps this draft grounded in what is publicly reproducible from chain data.\n\nSo the next theme after this one is simple:\n\nWhat the chain can show today, and what the paper says, would further strengthen the evaluation if it can be published safely.",
  "title": "PoBU, measured on a running chain",
  "updatedAt": "2026-03-03T18:20:38.327Z"
}