{
  "handle": "MetaPrecision",
  "gloss": "How comparable all attempts are under one shared yardstick, which is what makes them rankable",
  "mechanism": "Precision relates attempts to one another rather than to the target. Meta-precision is the same idea one level up: how far the conditions and measurement of every attempt have been standardised so that all attempts can be placed on one scale and ranked. An archery contest achieves high meta-precision by giving every archer the same distance, the same target and the same units. Mixing metric and imperial targets, or allowing each attempt its own target, lowers meta-precision because attempts become incommensurable. High meta-precision is what makes benchmarking, fair tests and optimisation by ranking possible.",
  "invariants": [
    "Meta-precision concerns comparability between attempts, not closeness to a target",
    "It rises with standardisation of conditions, targets and units",
    "Attempts measured against different targets or in different codes are incommensurable"
  ],
  "failure_modes": [
    "Mistaking high meta-precision (everything rankable) for success at the real goal",
    "Forcing a single yardstick onto a problem that is better solved by many incommensurable means"
  ],
  "_meta": {
    "path": [
      "Mind",
      "Reasoning"
    ],
    "ring": 2,
    "tier": 2,
    "source": "Adam Timlett, On the Origin of Risk (2025), Ch 14 p.98-103"
  },
  "sema_id": "sema:MetaPrecision#mh:SHA-256:63f292ef7c5542a9fa0f90bf974f6664c04dce490532f5bf7a3d107672282074",
  "sema_ref": "MetaPrecision#63f2",
  "sema_stub": "63f2",
  "sema_layer": "Mind",
  "sema_category": "Reasoning",
  "used_by": {
    "references": {
      "NaturalProblem": "sema:NaturalProblem#mh:SHA-256:ea59543ebcccb8df82bcc1a5e3dace47428a178230154ac53345b0e39daae102"
    }
  }
}