{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:HLURFTNVMPXVPUHVLI7QV3232H","short_pith_number":"pith:HLURFTNV","canonical_record":{"source":{"id":"2502.10985","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-16T04:07:33Z","cross_cats_sorted":["cs.AI","stat.ME","stat.ML"],"title_canon_sha256":"d7af3541565486f85e9a38120f88553cda91d3bed32af978185b51c2401991ae","abstract_canon_sha256":"6e7db8104e5f0b0c3316e01fe5ac18f72ed41c0a7f78a2fd1b6c17a872b0c52d"},"schema_version":"1.0"},"canonical_sha256":"3ae912cdb563ef57d0f55a3f0aef5bd1f1f53dad3872d27eb30ecb2100a43ca6","source":{"kind":"arxiv","id":"2502.10985","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.10985","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"arxiv_version","alias_value":"2502.10985v1","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.10985","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_12","alias_value":"HLURFTNVMPXV","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_16","alias_value":"HLURFTNVMPXVPUHV","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_8","alias_value":"HLURFTNV","created_at":"2026-07-05T10:15:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:HLURFTNVMPXVPUHVLI7QV3232H","target":"record","payload":{"canonical_record":{"source":{"id":"2502.10985","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-16T04:07:33Z","cross_cats_sorted":["cs.AI","stat.ME","stat.ML"],"title_canon_sha256":"d7af3541565486f85e9a38120f88553cda91d3bed32af978185b51c2401991ae","abstract_canon_sha256":"6e7db8104e5f0b0c3316e01fe5ac18f72ed41c0a7f78a2fd1b6c17a872b0c52d"},"schema_version":"1.0"},"canonical_sha256":"3ae912cdb563ef57d0f55a3f0aef5bd1f1f53dad3872d27eb30ecb2100a43ca6","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:02.091191Z","signature_b64":"JlO8SQAFL5a8AmyjPyf1lKCyVRdI9PmZppiS60AP51Cw6agf+qb2+cM6zGnkduNTnv7eXpaR7u7GPvPF+ljeDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ae912cdb563ef57d0f55a3f0aef5bd1f1f53dad3872d27eb30ecb2100a43ca6","last_reissued_at":"2026-07-05T10:15:02.090720Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:02.090720Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.10985","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:15:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zYCPWtvMAbA9g827rOJgzw9mMakRtVePE58PxGxsZz6+dOt7HTw3hcPuXd35Ya6xy7EAjJGZLjyLxit1kxeCAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T11:35:47.278409Z"},"content_sha256":"d3745f084d6e3dd85c2a4ec12aa3175bf09412847fe3779a325313fc88a07762","schema_version":"1.0","event_id":"sha256:d3745f084d6e3dd85c2a4ec12aa3175bf09412847fe3779a325313fc88a07762"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:HLURFTNVMPXVPUHVLI7QV3232H","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Is Elo Rating Reliable? A Study Under Model Misspecification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ME","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chi Jin, Shange Tang, Yuanhao Wang","submitted_at":"2025-02-16T04:07:33Z","abstract_excerpt":"Elo rating, widely used for skill assessment across diverse domains ranging from competitive games to large language models, is often understood as an incremental update algorithm for estimating a stationary Bradley-Terry (BT) model. However, our empirical analysis of practical matching datasets reveals two surprising findings: (1) Most games deviate significantly from the assumptions of the BT model and stationarity, raising questions on the reliability of Elo. (2) Despite these deviations, Elo frequently outperforms more complex rating systems, such as mElo and pairwise models, which are spe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.10985","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.10985/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:15:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fpwBu4Lkq7xA+A4OlkD8YPagexK8/PSuir6cEN+Kl/+yoUAYGmvie6tF0J1zJHEDb24xr71Ka1ahlJJdobPvAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T11:35:47.279028Z"},"content_sha256":"3be52a3730f988816bcce16a6f603faa9f19ffd87740ea34a0a824e286230aee","schema_version":"1.0","event_id":"sha256:3be52a3730f988816bcce16a6f603faa9f19ffd87740ea34a0a824e286230aee"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HLURFTNVMPXVPUHVLI7QV3232H/bundle.json","state_url":"https://pith.science/pith/HLURFTNVMPXVPUHVLI7QV3232H/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HLURFTNVMPXVPUHVLI7QV3232H/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-10T11:35:47Z","links":{"resolver":"https://pith.science/pith/HLURFTNVMPXVPUHVLI7QV3232H","bundle":"https://pith.science/pith/HLURFTNVMPXVPUHVLI7QV3232H/bundle.json","state":"https://pith.science/pith/HLURFTNVMPXVPUHVLI7QV3232H/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HLURFTNVMPXVPUHVLI7QV3232H/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:HLURFTNVMPXVPUHVLI7QV3232H","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6e7db8104e5f0b0c3316e01fe5ac18f72ed41c0a7f78a2fd1b6c17a872b0c52d","cross_cats_sorted":["cs.AI","stat.ME","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-16T04:07:33Z","title_canon_sha256":"d7af3541565486f85e9a38120f88553cda91d3bed32af978185b51c2401991ae"},"schema_version":"1.0","source":{"id":"2502.10985","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.10985","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"arxiv_version","alias_value":"2502.10985v1","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.10985","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_12","alias_value":"HLURFTNVMPXV","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_16","alias_value":"HLURFTNVMPXVPUHV","created_at":"2026-07-05T10:15:02Z"},{"alias_kind":"pith_short_8","alias_value":"HLURFTNV","created_at":"2026-07-05T10:15:02Z"}],"graph_snapshots":[{"event_id":"sha256:3be52a3730f988816bcce16a6f603faa9f19ffd87740ea34a0a824e286230aee","target":"graph","created_at":"2026-07-05T10:15:02Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.10985/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Elo rating, widely used for skill assessment across diverse domains ranging from competitive games to large language models, is often understood as an incremental update algorithm for estimating a stationary Bradley-Terry (BT) model. However, our empirical analysis of practical matching datasets reveals two surprising findings: (1) Most games deviate significantly from the assumptions of the BT model and stationarity, raising questions on the reliability of Elo. (2) Despite these deviations, Elo frequently outperforms more complex rating systems, such as mElo and pairwise models, which are spe","authors_text":"Chi Jin, Shange Tang, Yuanhao Wang","cross_cats":["cs.AI","stat.ME","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-16T04:07:33Z","title":"Is Elo Rating Reliable? A Study Under Model Misspecification"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.10985","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d3745f084d6e3dd85c2a4ec12aa3175bf09412847fe3779a325313fc88a07762","target":"record","created_at":"2026-07-05T10:15:02Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6e7db8104e5f0b0c3316e01fe5ac18f72ed41c0a7f78a2fd1b6c17a872b0c52d","cross_cats_sorted":["cs.AI","stat.ME","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-16T04:07:33Z","title_canon_sha256":"d7af3541565486f85e9a38120f88553cda91d3bed32af978185b51c2401991ae"},"schema_version":"1.0","source":{"id":"2502.10985","kind":"arxiv","version":1}},"canonical_sha256":"3ae912cdb563ef57d0f55a3f0aef5bd1f1f53dad3872d27eb30ecb2100a43ca6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3ae912cdb563ef57d0f55a3f0aef5bd1f1f53dad3872d27eb30ecb2100a43ca6","first_computed_at":"2026-07-05T10:15:02.090720Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:15:02.090720Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"JlO8SQAFL5a8AmyjPyf1lKCyVRdI9PmZppiS60AP51Cw6agf+qb2+cM6zGnkduNTnv7eXpaR7u7GPvPF+ljeDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:15:02.091191Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.10985","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d3745f084d6e3dd85c2a4ec12aa3175bf09412847fe3779a325313fc88a07762","sha256:3be52a3730f988816bcce16a6f603faa9f19ffd87740ea34a0a824e286230aee"],"state_sha256":"fb304121f0f2f20b792b580dc1dee082ec1a3d9efbef71548bb53226c55eee4a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rj9SGHCOrDf9WqHiweGo9CGG0nlP0R4iOY/2slBtWvU0YQf1KuYi8RGYbfpvRX3X4AglAIA6XokFTBVEN7QdCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T11:35:47.284088Z","bundle_sha256":"8e41e329b7a091c28e8573c7e41345e2c00549a438edd60adfc63d9a919ed30e"}}