{"id":"W2982119900","doi":"10.48550/arxiv.1910.12622","title":"Trouble with the Curve: Predicting Future MLB Players Using Scouting Reports","year":2019,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Sports Analytics and Performance","field":"Economics, Econometrics and Finance","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"League; Minor (academic); Valuation (finance); Computer science; Data science; Artificial intelligence; Operations research; Machine learning; World Wide Web; Engineering; Political science; Business; Accounting","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009808962,0.001135994,0.0003183563,0.002793566,0.0002876222,0.001352457,0.0007757524,0.0009544193,0.004274665],"category_scores_gemma":[0.004446999,0.0001797359,0.0004404902,0.001634765,0.0002389535,0.00100279,0.0009018853,0.00107782,0.004984197],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006628762,"about_ca_system_score_gemma":0.0005751926,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02608468,"about_ca_topic_score_gemma":0.04641771,"domain_scores_codex":[0.9996189,0.00008499462,0.00002164996,0.0001086189,0.00008627168,0.00007945075],"domain_scores_gemma":[0.9985299,0.0004291898,0.0002478229,0.0001451934,0.0002406155,0.0004072562],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.001395807,0.0007562362,0.6648917,0.0001614908,0.0001766416,0.0004599172,0.0003639237,0.03026722,0.001422481,0.001595649,0.09386566,0.2046434],"study_design_scores_gemma":[0.0001277577,0.0004179817,0.3713077,0.0002457376,0.00008125073,0.0003556726,0.001560377,0.575325,0.003659832,0.004433467,0.04239838,0.00008697765],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9204189,0.001720217,0.01050155,0.001540197,0.0003745962,0.0001193661,0.05113041,0.002061677,0.0121331],"genre_scores_gemma":[0.8843267,0.000662348,0.009617293,0.0001289317,0.0001875853,0.00006300498,0.09450632,0.0001365979,0.01037131],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.02608468,"threshold_uncertainty_score":0.0518657,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07420720122022083,"score_gpt":0.1721069129775375,"score_spread":0.09789971175731667,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}