{
  "_comment": "Judging modes the statement text cannot express, for the ICPC WF problems of this dataset (keys <year>/<letter>). Time limits and float tolerances otherwise come from statement.txt; each problem's judge.json has the resolved values. `abs_tolerance` + `rel_tolerance` together mean 'within either bound'. Validators follow the Kattis output-validator protocol (argv: input answer feedback_dir [flags], team output on stdin; exit 42 accept, 43 reject); interactors are wired to the submission over pipes.",
  "problems": {
    "2011/K": {
      "mode": "float",
      "tolerance": 0.010001,
      "tolerance_kind": "float_absolute_tolerance",
      "slug": "K-trash-removal",
      "inventory_kind": "float",
      "why": "statement: 'Answers within 1/100 of the correct rounded answer will be accepted' -- no 'error of 10^-k' wording, so the parser misses it; x1.0001 because 2.41-2.40 > 0.01 in floating point"
    },
    "2012/B": {
      "mode": "float",
      "tolerance": 0.010001,
      "tolerance_kind": "float_absolute_tolerance",
      "slug": "B-curvy-little-bottles",
      "inventory_kind": "float",
      "why": "statement: volumes and marks 'accurate to two decimal places'; one unit in the last printed place, x1.0001 so floating-point noise (2.41-2.40 > 0.01) is not a rejection"
    },
    "2012/G": {
      "mode": "float",
      "tolerance": 0.00010001,
      "tolerance_kind": "float_absolute_tolerance",
      "slug": "G-minimum-cost-flow",
      "inventory_kind": "float",
      "why": "statement: cost 'accurate to four decimal places'; one unit in the last printed place, x1.0001 so floating-point noise (2.41-2.40 > 0.01) is not a rejection"
    },
    "2012/H": {
      "mode": "float",
      "tolerance": 0.010001,
      "tolerance_kind": "float_absolute_tolerance",
      "slug": "H-room-service",
      "inventory_kind": "float",
      "why": "statement: route length 'accurate to two decimal places'; one unit in the last printed place, x1.0001 so floating-point noise (2.41-2.40 > 0.01) is not a rejection"
    },
    "2012/J": {
      "mode": "float",
      "tolerance": 0.0010001,
      "tolerance_kind": "float_absolute_tolerance",
      "slug": "J-shortest-flight-path",
      "inventory_kind": "float",
      "why": "statement: length 'accurate to three decimal places'; one unit in the last printed place, x1.0001 so floating-point noise (2.41-2.40 > 0.01) is not a rejection"
    },
    "2014/F": {
      "mode": "float",
      "abs_tolerance": 0.001,
      "rel_tolerance": 1e-05,
      "slug": "F-messenger",
      "inventory_kind": "float",
      "why": "statement: 'absolute error of at most 10^-3 or a relative error of at most 10^-5'; the parser keeps only the absolute part, 400x too strict on answers near 42425"
    },
    "2019/A": {
      "mode": "custom",
      "why": "any pair of row orderings meeting the price and height constraints is accepted",
      "slug": "A-azulejos",
      "inventory_kind": "multi-answer"
    },
    "2019/C": {
      "mode": "custom",
      "why": "any setup from which the moves are legal is accepted",
      "slug": "C-checks-post-facto",
      "inventory_kind": "multi-answer"
    },
    "2020/H": {
      "mode": "interactive",
      "why": "QC QC; the interactor replays the fixed report matrix, as testing_tool.py does",
      "slug": "H-qc-qc",
      "inventory_kind": "INTERACTIVE"
    },
    "2020/N": {
      "mode": "custom",
      "why": "any vector whose distances match within 1e-5 is accepted; underdetermined when n <= d",
      "slug": "N-whats-our-vector-victor",
      "inventory_kind": "float+multi"
    },
    "2021/H": {
      "mode": "custom",
      "why": "any permutation whose concatenation is balanced is accepted",
      "slug": "H-prehistoric-programs",
      "inventory_kind": "multi-answer"
    },
    "2022/U": {
      "mode": "custom",
      "why": "any loop of maximal length is accepted",
      "slug": "U-toy-train-tracks",
      "inventory_kind": "multi-answer"
    },
    "2022/Z": {
      "mode": "custom",
      "why": "any lever set producing the given configuration multiset is accepted",
      "slug": "Z-archaeological-recovery",
      "inventory_kind": "multi-answer"
    },
    "2023/A": {
      "mode": "interactive",
      "why": "Riddle of the Sphinx (same problem as 2022/W)",
      "slug": "A-riddle-of-the-sphinx",
      "inventory_kind": "INTERACTIVE"
    },
    "2023/H": {
      "mode": "custom",
      "why": "any valid sleep schedule is accepted; the period count is not minimised",
      "slug": "H-jet-lag",
      "inventory_kind": "multi-answer"
    },
    "2024/A": {
      "mode": "custom",
      "why": "any split giving every sponsor >= 1/n of their area is accepted",
      "slug": "A-billboards",
      "inventory_kind": "multi-answer"
    },
    "2024/D": {
      "mode": "custom",
      "why": "any maximum pairing is accepted",
      "slug": "D-doubles-horseback-wrestling",
      "inventory_kind": "multi-answer"
    },
    "2024/F": {
      "mode": "custom",
      "why": "any optimal team split is accepted",
      "slug": "F-friendly-rivalry",
      "inventory_kind": "float+multi"
    },
    "2024/L": {
      "mode": "interactive",
      "why": "Where Am I Now -- probe the grid, then state the position",
      "slug": "L-where-am-i-now",
      "inventory_kind": "INTERACTIVE"
    },
    "2025/B": {
      "mode": "custom",
      "why": "no .ans shipped at all -- 'output any valid winning first move'",
      "slug": "B-blackboard-game",
      "inventory_kind": "exact-diff"
    },
    "2025/I": {
      "mode": "interactive",
      "why": "Slot Machine -- pull levers, then state the arrangement",
      "slug": "I-slot-machine",
      "inventory_kind": "INTERACTIVE"
    },
    "2025/J": {
      "mode": "custom",
      "why": "any ordering of the cups reaching height h is accepted",
      "slug": "J-stacking-cups",
      "inventory_kind": "multi-answer"
    }
  }
}