{
  "schema": "windtunnel-reconciliation.v1",
  "reviewed_on": "2026-10-07",
  "measurement_period": "July 2026",
  "cohort": {
    "outputs": 192,
    "prompt_arms": 4,
    "lap_counts": [
      1,
      2,
      3,
      4
    ],
    "outputs_per_arm_lap": 12,
    "outputs_per_arm": 48,
    "scores_per_output": 3
  },
  "method": {
    "mean": "score_sum / count",
    "display": "Two decimal places, round half up. Sums and counts retain the exact means.",
    "comparison": "The same retained outputs were scored under three rubrics; no new model output was generated for this reconciliation.",
    "critic_model": "qwen3-coder:30b",
    "scoring_model": "qwen3-coder:30b"
  },
  "sources": [
    {
      "role": "original_outputs",
      "sha256": "38517198a430bd3ff7812860d19b390d4718aa77fd92396374b9d892667103f8",
      "access": "retained privately"
    },
    {
      "role": "original_rejudge_scores",
      "sha256": "34f6485051cf7bc45d84708081d7c0b027f746ad53098fee8505aeeca0e6a02d",
      "access": "retained privately"
    },
    {
      "role": "later_repeat_scores",
      "sha256": "da45e158520ae8a18532b587af4340d977d04ef6aed27e3dd9877975d5d5523d",
      "access": "retained privately"
    }
  ],
  "aggregates": [
    {
      "arm": "baseline",
      "laps": 1,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1022,
      "mean_display": "85.17"
    },
    {
      "arm": "baseline",
      "laps": 1,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "baseline",
      "laps": 1,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "baseline",
      "laps": 2,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1011,
      "mean_display": "84.25"
    },
    {
      "arm": "baseline",
      "laps": 2,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1075,
      "mean_display": "89.58"
    },
    {
      "arm": "baseline",
      "laps": 2,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1049,
      "mean_display": "87.42"
    },
    {
      "arm": "baseline",
      "laps": 3,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1012,
      "mean_display": "84.33"
    },
    {
      "arm": "baseline",
      "laps": 3,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1090,
      "mean_display": "90.83"
    },
    {
      "arm": "baseline",
      "laps": 3,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1069,
      "mean_display": "89.08"
    },
    {
      "arm": "baseline",
      "laps": 4,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 894,
      "mean_display": "74.50"
    },
    {
      "arm": "baseline",
      "laps": 4,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1086,
      "mean_display": "90.50"
    },
    {
      "arm": "baseline",
      "laps": 4,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1038,
      "mean_display": "86.50"
    },
    {
      "arm": "baseline",
      "laps": "all",
      "rubric": "directness-default",
      "count": 48,
      "score_sum": 3939,
      "mean_display": "82.06"
    },
    {
      "arm": "baseline",
      "laps": "all",
      "rubric": "completeness",
      "count": 48,
      "score_sum": 4334,
      "mean_display": "90.29"
    },
    {
      "arm": "baseline",
      "laps": "all",
      "rubric": "neutral",
      "count": 48,
      "score_sum": 4239,
      "mean_display": "88.31"
    },
    {
      "arm": "concise-author",
      "laps": 1,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1097,
      "mean_display": "91.42"
    },
    {
      "arm": "concise-author",
      "laps": 1,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1069,
      "mean_display": "89.08"
    },
    {
      "arm": "concise-author",
      "laps": 1,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1097,
      "mean_display": "91.42"
    },
    {
      "arm": "concise-author",
      "laps": 2,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1090,
      "mean_display": "90.83"
    },
    {
      "arm": "concise-author",
      "laps": 2,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1090,
      "mean_display": "90.83"
    },
    {
      "arm": "concise-author",
      "laps": 2,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1097,
      "mean_display": "91.42"
    },
    {
      "arm": "concise-author",
      "laps": 3,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1056,
      "mean_display": "88.00"
    },
    {
      "arm": "concise-author",
      "laps": 3,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1099,
      "mean_display": "91.58"
    },
    {
      "arm": "concise-author",
      "laps": 3,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1093,
      "mean_display": "91.08"
    },
    {
      "arm": "concise-author",
      "laps": 4,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "concise-author",
      "laps": 4,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "concise-author",
      "laps": 4,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1100,
      "mean_display": "91.67"
    },
    {
      "arm": "concise-author",
      "laps": "all",
      "rubric": "directness-default",
      "count": 48,
      "score_sum": 4326,
      "mean_display": "90.13"
    },
    {
      "arm": "concise-author",
      "laps": "all",
      "rubric": "completeness",
      "count": 48,
      "score_sum": 4341,
      "mean_display": "90.44"
    },
    {
      "arm": "concise-author",
      "laps": "all",
      "rubric": "neutral",
      "count": 48,
      "score_sum": 4387,
      "mean_display": "91.40"
    },
    {
      "arm": "minimalist-critic",
      "laps": 1,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1065,
      "mean_display": "88.75"
    },
    {
      "arm": "minimalist-critic",
      "laps": 1,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "minimalist-critic",
      "laps": 1,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1090,
      "mean_display": "90.83"
    },
    {
      "arm": "minimalist-critic",
      "laps": 2,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1042,
      "mean_display": "86.83"
    },
    {
      "arm": "minimalist-critic",
      "laps": 2,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1069,
      "mean_display": "89.08"
    },
    {
      "arm": "minimalist-critic",
      "laps": 2,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "minimalist-critic",
      "laps": 3,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1035,
      "mean_display": "86.25"
    },
    {
      "arm": "minimalist-critic",
      "laps": 3,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1062,
      "mean_display": "88.50"
    },
    {
      "arm": "minimalist-critic",
      "laps": 3,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1069,
      "mean_display": "89.08"
    },
    {
      "arm": "minimalist-critic",
      "laps": 4,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1038,
      "mean_display": "86.50"
    },
    {
      "arm": "minimalist-critic",
      "laps": 4,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "minimalist-critic",
      "laps": 4,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1069,
      "mean_display": "89.08"
    },
    {
      "arm": "minimalist-critic",
      "laps": "all",
      "rubric": "directness-default",
      "count": 48,
      "score_sum": 4180,
      "mean_display": "87.08"
    },
    {
      "arm": "minimalist-critic",
      "laps": "all",
      "rubric": "completeness",
      "count": 48,
      "score_sum": 4283,
      "mean_display": "89.23"
    },
    {
      "arm": "minimalist-critic",
      "laps": "all",
      "rubric": "neutral",
      "count": 48,
      "score_sum": 4304,
      "mean_display": "89.67"
    },
    {
      "arm": "thorough-critic",
      "laps": 1,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 1014,
      "mean_display": "84.50"
    },
    {
      "arm": "thorough-critic",
      "laps": 1,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "thorough-critic",
      "laps": 1,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "thorough-critic",
      "laps": 2,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 964,
      "mean_display": "80.33"
    },
    {
      "arm": "thorough-critic",
      "laps": 2,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "thorough-critic",
      "laps": 2,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1076,
      "mean_display": "89.67"
    },
    {
      "arm": "thorough-critic",
      "laps": 3,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 951,
      "mean_display": "79.25"
    },
    {
      "arm": "thorough-critic",
      "laps": 3,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1083,
      "mean_display": "90.25"
    },
    {
      "arm": "thorough-critic",
      "laps": 3,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1042,
      "mean_display": "86.83"
    },
    {
      "arm": "thorough-critic",
      "laps": 4,
      "rubric": "directness-default",
      "count": 12,
      "score_sum": 921,
      "mean_display": "76.75"
    },
    {
      "arm": "thorough-critic",
      "laps": 4,
      "rubric": "completeness",
      "count": 12,
      "score_sum": 1082,
      "mean_display": "90.17"
    },
    {
      "arm": "thorough-critic",
      "laps": 4,
      "rubric": "neutral",
      "count": 12,
      "score_sum": 1049,
      "mean_display": "87.42"
    },
    {
      "arm": "thorough-critic",
      "laps": "all",
      "rubric": "directness-default",
      "count": 48,
      "score_sum": 3850,
      "mean_display": "80.21"
    },
    {
      "arm": "thorough-critic",
      "laps": "all",
      "rubric": "completeness",
      "count": 48,
      "score_sum": 4324,
      "mean_display": "90.08"
    },
    {
      "arm": "thorough-critic",
      "laps": "all",
      "rubric": "neutral",
      "count": 48,
      "score_sum": 4243,
      "mean_display": "88.40"
    }
  ],
  "repeat_scoring": {
    "scope": "Separate later neutral-rubric repeat-scoring cohort; not pooled into the original rejudge means.",
    "outputs": 192,
    "ratings_per_output": 6,
    "ratings": 1152,
    "mean_within_cell_population_sd": 0.39969422221655215,
    "mean_within_cell_sample_sd": 0.4378430832249763,
    "by_arm": [
      {
        "arm": "baseline",
        "count": 288,
        "score_sum": 25375,
        "mean_display": "88.11"
      },
      {
        "arm": "concise-author",
        "count": 288,
        "score_sum": 26252,
        "mean_display": "91.15"
      },
      {
        "arm": "minimalist-critic",
        "count": 288,
        "score_sum": 25845,
        "mean_display": "89.74"
      },
      {
        "arm": "thorough-critic",
        "count": 288,
        "score_sum": 25309,
        "mean_display": "87.88"
      }
    ]
  },
  "rubric_sensitivity": {
    "definition": "For each original output, highest minus lowest score across the three rubrics; then mean across outputs.",
    "mean_within_output_score_range": 6.505208333333333,
    "maximum_within_output_score_range": 47.0
  },
  "limits": [
    "Model-graded scores, not independent human assessments or general model reliability.",
    "The critic and scorer used the same model; varying the scoring rubric is not independent-model judging.",
    "Only counts, sums and derived statistics are published. Cell identifiers, prompts, model answers and operational paths are withheld.",
    "Source hashes identify retained bytes; they do not authenticate the underlying measurements for a reader without those records."
  ]
}