{
  "id": "goodhart-resistance",
  "type": "entity",
  "name": "Goodhart resistance in agent metrics",
  "summary": "Designing agent evaluation so that optimizing the reported number does not destroy what the number means. Known countermeasures: separate the metric you optimize from the metric you report, measure consistency across repeated trials rather than best-of-N, and audit with a panel of honesty figures instead of a single score.",
  "locale": "en",
  "tags": [
    "goodhart",
    "metrics",
    "evaluation",
    "gaming"
  ],
  "relations": [
    {
      "rel": "threatened_by",
      "target": "llm-as-judge"
    },
    {
      "rel": "defended_by",
      "target": "opaque-rotating-test-sets"
    }
  ],
  "questions": [
    "How do I stop my team (or my agent) from gaming the eval?",
    "Why is a single quality score dangerous?"
  ],
  "claims": [
    {
      "id": "c1",
      "text": "Goodhart's Law is not one failure but at least four distinct mechanisms — regressional, extremal, causal and adversarial — each requiring different defenses.",
      "sources": [
        "goodhart-variants"
      ],
      "confidence": 0.9
    },
    {
      "id": "c2",
      "text": "Single-run success overstates reliability: on tau-bench, state-of-the-art function-calling agents pass under 50% of tasks once, and pass^8 falls under 25% in the retail domain, which is why pass^k exists as a consistency metric.",
      "sources": [
        "tau-bench"
      ],
      "confidence": 0.9
    },
    {
      "id": "c3",
      "text": "The Citarium methodology separates CAS-I (internal score) from CAS-E (external score) precisely so the optimized metric and the reported metric are not the same object.",
      "sources": [
        "citarium-methodology-brief"
      ],
      "confidence": 0.85
    },
    {
      "id": "c4",
      "text": "The Citarium audit command reports a panel of honesty figures — evidence-tier mix vs quota, unused sources, moat count, staleness — rather than one aggregatable score, making the audit itself harder to Goodhart.",
      "sources": [
        "citarium-cli-audit"
      ],
      "confidence": 0.9
    }
  ],
  "takeaways": [],
  "faqs": [],
  "evidence_tier": "secondary",
  "evidence": {
    "level": "production",
    "source_types": [
      "production_system",
      "personal_experience",
      "paper"
    ]
  },
  "moat_flag": true,
  "winning_edge": "Pairs the academic taxonomy with two working countermeasures from a production framework (CAS-I/CAS-E separation, honesty-figure audits), where the usual coverage stops at \"beware Goodhart\".",
  "confidence": 0.85,
  "last_verified": "2026-08-08",
  "canonical_url": "https://agentreliability.dev/k/goodhart-resistance",
  "api_url": "https://agentreliability.dev/api/k/goodhart-resistance.json",
  "jsonld": {
    "@context": "https://schema.org",
    "name": "Goodhart resistance in agent metrics",
    "description": "Designing agent evaluation so that optimizing the reported number does not destroy what the number means. Known countermeasures: separate the metric you optimize from the metric you report, measure consistency across repeated trials rather than best-of-N, and audit with a panel of honesty figures instead of a single score.",
    "url": "https://agentreliability.dev/k/goodhart-resistance",
    "license": "https://spdx.org/licenses/CC-BY-4.0.html",
    "dateModified": "2026-08-08",
    "citation": [
      {
        "@type": "CreativeWork",
        "name": "Citarium audit command source (@citarium/cli v0.1.0, commands.ts)"
      },
      {
        "@type": "CreativeWork",
        "name": "agent-reliability editorial brief and blueprint (Gate 1 approved, 2026-08-08)"
      },
      {
        "@type": "CreativeWork",
        "name": "Categorizing Variants of Goodhart's Law",
        "url": "https://arxiv.org/abs/1803.04585"
      },
      {
        "@type": "CreativeWork",
        "name": "tau-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains",
        "url": "https://arxiv.org/abs/2406.12045"
      }
    ],
    "author": {
      "@type": "Person",
      "name": "Santiago Santa María Morales",
      "jobTitle": "practitioner — harness engineering and agent evaluation in production"
    },
    "@type": "Article",
    "headline": "Goodhart resistance in agent metrics"
  }
}