{
  "sourceRun": "runs/fly-aggregate-002",
  "displayNames": {
    "meanown-cap": "Flomb"
  },
  "trainingGamesPerSeed": 20000,
  "trainingSeeds": [
    10,
    11,
    12,
    13,
    14,
    15,
    16,
    17,
    18,
    19
  ],
  "evaluationSets": [
    {
      "size": 7,
      "mines": 7,
      "gamesPerSeed": 250
    },
    {
      "size": 9,
      "mines": 12,
      "gamesPerSeed": 500
    },
    {
      "size": 16,
      "mines": 40,
      "gamesPerSeed": 150
    }
  ],
  "metric": "percentage of episodes won",
  "interval": "95% Student t confidence interval across 10 training seeds (df=9)",
  "reference": "local hand-coded heuristic on the same layouts, tie-breaking seed 0; no training-seed interval",
  "sourceFiles": {
    "benchmark-summary.json": "75b2ade27c63e292e2eac0d1570a19665ee4a08e06498772829a5e9575336acc",
    "benchmark-manifest.json": "3eddd57e05ac450087457a7dd5cc4ac7065a2f57cc2b1c7ff0913a9d5d6c45b3",
    "benchmark-protocolo.md": "263d27ed34cf1c4b6d52740362b8d4b880c3d7925a598e8c814967943cec62ae"
  },
  "validation": "reported per-seed statistics checked; original per-game traces unavailable in this checkout",
  "results": {
    "7": {
      "sum-cap": {
        "mean": 84.3,
        "sd": 2.0,
        "ci95": [
          82.9,
          85.7
        ],
        "per_seed": [
          87.2,
          82.4,
          84.0,
          82.0,
          85.6,
          81.6,
          86.0,
          83.2,
          84.4,
          86.4
        ]
      },
      "mean-cap": {
        "mean": 87.4,
        "sd": 1.3,
        "ci95": [
          86.5,
          88.4
        ],
        "per_seed": [
          88.8,
          85.2,
          88.8,
          88.4,
          88.8,
          85.6,
          87.2,
          87.2,
          86.4,
          88.0
        ]
      },
      "meanown-cap": {
        "mean": 88.3,
        "sd": 3.7,
        "ci95": [
          85.7,
          91.0
        ],
        "per_seed": [
          84.4,
          87.6,
          90.4,
          94.0,
          85.2,
          90.0,
          91.6,
          81.6,
          90.4,
          88.0
        ]
      },
      "mean-cap minus sum-cap": {
        "mean": 3.2,
        "sd": 1.7,
        "ci95": [
          2.0,
          4.3
        ],
        "per_seed": [
          1.6,
          2.8,
          4.8,
          6.4,
          3.2,
          4.0,
          1.2,
          4.0,
          2.0,
          1.6
        ],
        "p": 0.0002
      },
      "meanown-cap minus sum-cap": {
        "mean": 4.0,
        "sd": 4.7,
        "ci95": [
          0.7,
          7.4
        ],
        "per_seed": [
          -2.8,
          5.2,
          6.4,
          12.0,
          -0.4,
          8.4,
          5.6,
          -1.6,
          6.0,
          1.6
        ],
        "p": 0.0239
      },
      "meanown-cap minus mean-cap": {
        "mean": 0.9,
        "sd": 4.1,
        "ci95": [
          -2.0,
          3.8
        ],
        "per_seed": [
          -4.4,
          2.4,
          1.6,
          5.6,
          -3.6,
          4.4,
          4.4,
          -5.6,
          4.0,
          0.0
        ],
        "p": 0.5131
      },
      "hand": 91.2,
      "safe_step_pct": {
        "sum-cap": 98.51,
        "mean-cap": 98.77,
        "meanown-cap": 98.85
      },
      "false_marks": {
        "sum-cap": 4,
        "mean-cap": 6,
        "meanown-cap": 0
      }
    },
    "9": {
      "sum-cap": {
        "mean": 76.3,
        "sd": 2.2,
        "ci95": [
          74.7,
          77.8
        ],
        "per_seed": [
          79.0,
          75.0,
          75.2,
          76.2,
          77.8,
          71.2,
          77.2,
          77.8,
          76.4,
          76.8
        ]
      },
      "mean-cap": {
        "mean": 81.3,
        "sd": 1.8,
        "ci95": [
          80.0,
          82.7
        ],
        "per_seed": [
          81.0,
          77.6,
          83.2,
          83.2,
          83.8,
          79.8,
          81.6,
          81.0,
          80.4,
          81.8
        ]
      },
      "meanown-cap": {
        "mean": 81.3,
        "sd": 3.5,
        "ci95": [
          78.8,
          83.8
        ],
        "per_seed": [
          74.6,
          80.8,
          83.8,
          82.8,
          80.2,
          84.0,
          86.2,
          76.6,
          81.6,
          82.6
        ]
      },
      "mean-cap minus sum-cap": {
        "mean": 5.1,
        "sd": 2.3,
        "ci95": [
          3.5,
          6.7
        ],
        "per_seed": [
          2.0,
          2.6,
          8.0,
          7.0,
          6.0,
          8.6,
          4.4,
          3.2,
          4.0,
          5.0
        ],
        "p": 0.0001
      },
      "meanown-cap minus sum-cap": {
        "mean": 5.1,
        "sd": 5.0,
        "ci95": [
          1.5,
          8.7
        ],
        "per_seed": [
          -4.4,
          5.8,
          8.6,
          6.6,
          2.4,
          12.8,
          9.0,
          -1.2,
          5.2,
          5.8
        ],
        "p": 0.0111
      },
      "meanown-cap minus mean-cap": {
        "mean": -0.0,
        "sd": 3.7,
        "ci95": [
          -2.7,
          2.6
        ],
        "per_seed": [
          -6.4,
          3.2,
          0.6,
          -0.4,
          -3.6,
          4.2,
          4.6,
          -4.4,
          1.2,
          0.8
        ],
        "p": 0.9868
      },
      "hand": 84.8,
      "safe_step_pct": {
        "sum-cap": 98.72,
        "mean-cap": 98.98,
        "meanown-cap": 98.98
      },
      "false_marks": {
        "sum-cap": 32,
        "mean-cap": 65,
        "meanown-cap": 0
      }
    },
    "16": {
      "sum-cap": {
        "mean": 44.3,
        "sd": 7.8,
        "ci95": [
          38.7,
          49.8
        ],
        "per_seed": [
          50.7,
          42.0,
          43.3,
          38.7,
          46.0,
          26.0,
          52.7,
          51.3,
          45.3,
          46.7
        ]
      },
      "mean-cap": {
        "mean": 51.8,
        "sd": 5.6,
        "ci95": [
          47.8,
          55.8
        ],
        "per_seed": [
          50.0,
          49.3,
          55.3,
          54.7,
          60.7,
          38.7,
          51.3,
          53.3,
          53.3,
          51.3
        ]
      },
      "meanown-cap": {
        "mean": 57.3,
        "sd": 7.3,
        "ci95": [
          52.1,
          62.5
        ],
        "per_seed": [
          42.0,
          58.7,
          66.0,
          60.0,
          53.3,
          62.7,
          58.0,
          49.3,
          58.7,
          64.0
        ]
      },
      "mean-cap minus sum-cap": {
        "mean": 7.5,
        "sd": 6.3,
        "ci95": [
          3.0,
          12.0
        ],
        "per_seed": [
          -0.7,
          7.3,
          12.0,
          16.0,
          14.7,
          12.7,
          -1.3,
          2.0,
          8.0,
          4.7
        ],
        "p": 0.0042
      },
      "meanown-cap minus sum-cap": {
        "mean": 13.0,
        "sd": 13.1,
        "ci95": [
          3.6,
          22.4
        ],
        "per_seed": [
          -8.7,
          16.7,
          22.7,
          21.3,
          7.3,
          36.7,
          5.3,
          -2.0,
          13.3,
          17.3
        ],
        "p": 0.0119
      },
      "meanown-cap minus mean-cap": {
        "mean": 5.5,
        "sd": 9.9,
        "ci95": [
          -1.6,
          12.5
        ],
        "per_seed": [
          -8.0,
          9.3,
          10.7,
          5.3,
          -7.3,
          24.0,
          6.7,
          -4.0,
          5.3,
          12.7
        ],
        "p": 0.1132
      },
      "hand": 61.3,
      "safe_step_pct": {
        "sum-cap": 99.12,
        "mean-cap": 99.24,
        "meanown-cap": 99.33
      },
      "false_marks": {
        "sum-cap": 34,
        "mean-cap": 74,
        "meanown-cap": 0
      }
    }
  }
}
