{"paper":{"title":"Robust Player-Conditional Champion Ranking for League of Legends: Style Similarity, Mastery Priors, and Archetype-Constrained Discovery","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A framework ranks League of Legends champions for specific players by blending population strength, style similarity, mastery experience, and archetype constraints.","cross_cats":["cs.LG"],"primary_cat":"stat.AP","authors_text":"Min Heo, Pranav Kadiyam, Prasun Panthi","submitted_at":"2026-05-18T12:52:16Z","abstract_excerpt":"Champion recommendation in multiplayer online battle arena games is usually framed informally as a problem of metagame strength, personal comfort, or global win rate. We formalize champion recommendation in League of Legends as an interpretable, player-conditional ranking problem under sparse, noisy, and non-stationary behavioral data. The proposed framework combines four information sources: a population-strength proxy, player-style similarity, direct and indirect mastery priors, and archetype-level guardrails. The method uses robust median/MAD normalization, logarithmic transforms for skewed"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"The proposed framework combines four information sources: a population-strength proxy, player-style similarity, direct and indirect mastery priors, and archetype-level guardrails to produce interpretable player-conditional champion rankings under sparse, noisy, and non-stationary behavioral data.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That robust median/MAD normalization, logarithmic transforms, recency-weighted style vectors, mastery-weighted champion-pool vectors, weighted cosine similarity, and k-means++ archetype clustering can be combined into stable, meaningful rankings without post-hoc tuning that undermines the claimed interpretability and robustness.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A modular framework for interpretable, player-conditional champion ranking in League of Legends is specified using style similarity, mastery priors, archetype guardrails, and robust normalization on behavioral data.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A framework ranks League of Legends champions for specific players by blending population strength, style similarity, mastery experience, and archetype constraints.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5e5701e16b3286e44b17b7c1f90e6d80d28c3b45c024a609ff1df8073bb16844"},"source":{"id":"2605.18338","kind":"arxiv","version":1},"verdict":{"id":"507bacc6-5309-4857-a839-43848db1e15a","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T23:54:24.812425Z","strongest_claim":"The proposed framework combines four information sources: a population-strength proxy, player-style similarity, direct and indirect mastery priors, and archetype-level guardrails to produce interpretable player-conditional champion rankings under sparse, noisy, and non-stationary behavioral data.","one_line_summary":"A modular framework for interpretable, player-conditional champion ranking in League of Legends is specified using style similarity, mastery priors, archetype guardrails, and robust normalization on behavioral data.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That robust median/MAD normalization, logarithmic transforms, recency-weighted style vectors, mastery-weighted champion-pool vectors, weighted cosine similarity, and k-means++ archetype clustering can be combined into stable, meaningful rankings without post-hoc tuning that undermines the claimed interpretability and robustness.","pith_extraction_headline":"A framework ranks League of Legends champions for specific players by blending population strength, style similarity, mastery experience, and archetype constraints."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.18338/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_compliance","ran_at":"2026-05-20T00:02:35.017994Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-20T00:01:20.428924Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T23:33:35.167545Z","status":"skipped","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T23:21:58.833592Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"15b04946f3971bffbab91f5e7ca0f520f5cb2c704a6271ec83afb3a32948a0c6"},"references":{"count":58,"sample":[{"doi":"","year":null,"title":"We formulate champion recommendation as a content-based, player-conditional ranking problem with decomposable utility components","work_id":"1bf57d7d-7801-4cd9-80c8-35c880e13fca","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"We define a robust preprocessing pipeline using me- dian/MAD normalization, sign correction for nega- tive metrics, clipping, and log transforms for skewed event counts","work_id":"ba2d2027-d314-49d1-b86d-442f5c4f93f1","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"We construct player representations from both recency-weighted match behavior and mastery- weighted champion-pool behavior","work_id":"66947359-794a-4d73-bb01-4abb7d765529","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"We introduce a weighted cosine fit score whose fea- ture weights depend on both population dispersion and player-specific behavioral salience","work_id":"e98e3fd2-ad82-49c5-936d-6294778fb4c4","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2026,"title":"Robust Player-Conditional Champion Ranking for League of Legends: Style Similarity, Mastery Priors, and Archetype-Constrained Discovery","work_id":"92793872-0b3f-4f84-9150-8429830ead0e","ref_index":5,"cited_arxiv_id":"2605.18338","is_internal_anchor":true}],"resolved_work":58,"snapshot_sha256":"205c1454b53a8714d05f4aa331dc50d572945e5e95ab789cf4bf5921961fca72","internal_anchors":1},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1e17c95aa1400ca6e9728ac04476f259dbebcdf8f9ae2438dc3bf26c8e3ffdbe"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}