{"abstract":"Rank fusion uses incomparable raw scores.","category":"Search retrieval semantics","checks":6,"contract":"Each input list contains unique [id,score] hits in ranked order. Fuse by sum(1/(offset+rank)) using one-based rank; return IDs descending by exact rational fused score, lexical ties. Offset is a positive integer.","contract_signature":"lists, offset","evaluation_group":"model-0cb1403e1ae55c23","failed_approach":"Reciprocal zero-based ranks shift every contribution and change cross-list tradeoffs.","family":"z-search-reciprocal-rank-fusion","id":"FA-11866","implementations":{"attempt":{"sha256":"7732f4e5fc7df32a05d752bd7db97e589081cfb30c562a02ee1c24240a29995c","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(lists, offset):\n    from fractions import Fraction\n    scores={}\n    for hits in lists:\n        for rank,(ident,score) in enumerate(hits): scores[ident]=scores.get(ident,0)+Fraction(1,offset+rank)\n    return sorted(scores,key=lambda ident:(-scores[ident],ident))\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('scale irrelevant', solve([[['a',N],['b',1000*N]],[['b',N]]],1),['b','a'])\ncheck('one versus two deep ranks', solve([[['a',100],['x',2],['b',1]],[['y',5],['z',4],['b',1]]],1),['a','b','y','x','z'])\ncheck('lexical ties', solve([[['z',N]],[['a',2*N]]],2),['a','z'])\ncheck('empty lists', solve([[],[]],N),[])\ncheck('single list rank order', solve([[['z',1],['a',100]]],N),['z','a'])\ncheck('same document two lists', solve([[['a',0]],[['a',0]]],N),['a'])\nprint(json.dumps({\"observations\": observations, \"passed\": all(x[\"passed\"] for x in observations)}, ensure_ascii=False))\nraise SystemExit(0 if all(x[\"passed\"] for x in observations) else 1)\n"},"broken":{"sha256":"49968d997efef2c4cf9608d328dc709bd307cdbde2a5802fd03dccaa0dd3b316","source":"\"\"\"Failure Map reference implementation. Python standard library only.\"\"\"\nimport json\n\nN = 1\nobservations = []\ndef solve(lists, offset):\n    scores={}\n    for hits in lists:\n        for ident,score in hits: scores[ident]=scores.get(ident,0)+score\n    return sorted(scores,key=lambda ident:(-scores[ident],ident))\ndef check(label, actual, expected):\n    observations.append({\"check\": label, \"actual\": actual, \"expected\": expected, \"passed\": actual == expected})\ncheck('scale irrelevant', solve([[['a',N],['b',1000*N]],[['b',N]]],1),['b','a'])\ncheck('one versus two deep ranks', solve([[['a',100],['x',2],['b',1]],[['y',5],['z',4],['b',1]]],1),['a','b','y','x','z'])\ncheck('lexical ties', solve([[['z',N]],[['a',2*N]]],2),['a','z'])\ncheck('empty lists', solve([[],[]],N),[])\ncheck('single list rank order', solve([[['z',1],['a',100]]],N),['z','a'])\ncheck('same document two lists', solve([[['a',0]],[['a',0]]],N),['a'])\nprint(json.dumps({\"observations\": observations, \"passed\": all(x[\"passed\"] for x in observations)}, ensure_ascii=False))\nraise SystemExit(0 if all(x[\"passed\"] for x in observations) else 1)\n"}},"limitations":"Inputs are already tokenized or scored; this model makes no claim about production engine performance or linguistic analysis. This reproducer isolates one failure mechanism. Results cover the supplied fixtures. Variants within a family share a test contract and should remain grouped when constructing evaluation splits. Related mechanisms with a shared evaluation_group must also remain together; these controlled models are not independent production incidents.","method":"Deterministic executable model with adversarial boundary fixtures.","provenance":{"created_by":"Failure Map","dependencies":"Python standard library","family":"z-search-reciprocal-rank-fusion","generated_at":"2026-09-29T14:38:51.719578+00:00","license":"CC0-1.0","python":"3.12.14","seed":1,"split":"open-access"},"relevance":"An offline deterministic retrieval model isolates this search contract from tokenization, storage, and network behavior.","root_cause":"Scores from independent retrievers are summed despite incompatible scales.","sha256":"b5d972334785f670fcc5293b7164b96b6eee030d7f919390c96725d471cc7eea","title":"Rank fusion uses incomparable raw scores · case 01","variant":1,"variant_policy":"Five numbered records share a model and may reuse boundary fixtures.","verified":true,"visibility":"public","verification":{"attempt":{"elapsed_ms":43.318,"exit_code":1,"observations":[{"actual":["b","a"],"check":"scale irrelevant","expected":["b","a"],"passed":true},{"actual":["a","y","b","x","z"],"check":"one versus two deep ranks","expected":["a","b","y","x","z"],"passed":false},{"actual":["a","z"],"check":"lexical ties","expected":["a","z"],"passed":true},{"actual":[],"check":"empty lists","expected":[],"passed":true},{"actual":["z","a"],"check":"single list rank order","expected":["z","a"],"passed":true},{"actual":["a"],"check":"same document two lists","expected":["a"],"passed":true}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"scale irrelevant\", \"actual\": [\"b\", \"a\"], \"expected\": [\"b\", \"a\"], \"passed\": true}, {\"check\": \"one versus two deep ranks\", \"actual\": [\"a\", \"y\", \"b\", \"x\", \"z\"], \"expected\": [\"a\", \"b\", \"y\", \"x\", \"z\"], \"passed\": false}, {\"check\": \"lexical ties\", \"actual\": [\"a\", \"z\"], \"expected\": [\"a\", \"z\"], \"passed\": true}, {\"check\": \"empty lists\", \"actual\": [], \"expected\": [], \"passed\": true}, {\"check\": \"single list rank order\", \"actual\": [\"z\", \"a\"], \"expected\": [\"z\", \"a\"], \"passed\": true}, {\"check\": \"same document two lists\", \"actual\": [\"a\"], \"expected\": [\"a\"], \"passed\": true}], \"passed\": false}\n"},"broken":{"elapsed_ms":41.422,"exit_code":1,"observations":[{"actual":["b","a"],"check":"scale irrelevant","expected":["b","a"],"passed":true},{"actual":["a","y","z","b","x"],"check":"one versus two deep ranks","expected":["a","b","y","x","z"],"passed":false},{"actual":["a","z"],"check":"lexical ties","expected":["a","z"],"passed":true},{"actual":[],"check":"empty lists","expected":[],"passed":true},{"actual":["a","z"],"check":"single list rank order","expected":["z","a"],"passed":false},{"actual":["a"],"check":"same document two lists","expected":["a"],"passed":true}],"passed":false,"stderr":"","stdout":"{\"observations\": [{\"check\": \"scale irrelevant\", \"actual\": [\"b\", \"a\"], \"expected\": [\"b\", \"a\"], \"passed\": true}, {\"check\": \"one versus two deep ranks\", \"actual\": [\"a\", \"y\", \"z\", \"b\", \"x\"], \"expected\": [\"a\", \"b\", \"y\", \"x\", \"z\"], \"passed\": false}, {\"check\": \"lexical ties\", \"actual\": [\"a\", \"z\"], \"expected\": [\"a\", \"z\"], \"passed\": true}, {\"check\": \"empty lists\", \"actual\": [], \"expected\": [], \"passed\": true}, {\"check\": \"single list rank order\", \"actual\": [\"a\", \"z\"], \"expected\": [\"z\", \"a\"], \"passed\": false}, {\"check\": \"same document two lists\", \"actual\": [\"a\"], \"expected\": [\"a\"], \"passed\": true}], \"passed\": false}\n"}},"member_only":{"stages":["fixed"],"fields":["implementations.fixed","verification.fixed","harness","repair"],"note":"The verified repair, its recorded checks, the repair description, and the scoring harness are available to members."}}