{"paper_id":"2311.04599","title":"Explainable artificial intelligence model for identifying Market Value in Professional Soccer Players","authors":["Chunyang Huang","Shaoliang Zhang"],"paper_date":"2023-11-08","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2311.04599v2","page_url":"https://thequant.space/flowcharts/explainable-artificial-intelligence-model-for-identifying-ma/","scores":{"empirical_rigor":7,"math_complexity":2.5,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes an advanced machine learning method that combines ensemble models with SHAP for interpretable prediction of soccer players' market values.","quote":"This study introduces an advanced machine learning method for predicting soccer players' market values, combining ensemble models and the Shapley Additive Explanations (SHAP) for interpretability.","summary":"This study introduces an advanced machine learning method for predicting soccer players’ market values, combining ensemble models and the Shapley Additive Explanations (SHAP) for interpretability. Utilizing data from about 12,000 players from Sofifa, the Boruta algorithm streamlined feature…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The data is split into 80% training/validation and 20% testing, with imputation and feature selection performed only on the training set to avoid data leakage.","type":"data","stated":"explicit","quote":"In the construction of the predictive model, the dataset underwent a randomized spliting process: 80% was allocated for training and model validation, while the remaining 20% was designated for testing purposes."},{"assumption":"Ensemble learning improves predictive accuracy by reducing variance and bias compared to single models.","type":"model","stated":"explicit","quote":"The rationale behind selecting ensemble modeling lies in its multiple benefits, which include diminishing variance and bias while concurrently enhancing the overall predictive accuracy of the model."},{"assumption":"Hyperparameters are tuned using 5‑fold cross‑validation and grid search to obtain unbiased performance estimates.","type":"model","stated":"explicit","quote":"To optimize the hyperparameters of each ensemble learning model, we employed a 5-fold cross-validation technique coupled with Grid Search."},{"assumption":"The Boruta algorithm can effectively identify the most relevant features for the market‑value prediction task.","type":"model","stated":"explicit","quote":"Utilization of the Boruta algorithm, implemented through the BorutaShap package in Python, allowed for a reduction in the feature set to 22."},{"assumption":"Player attributes in skills, fitness, and cognitive areas are the primary drivers of market value.","type":"market","stated":"explicit","quote":"Player attributes in skills, fitness, and cognitive areas significantly influenced market value."}]},"datasets":{"status":"available","items":[{"name":"Sofifa player dataset","provider":"Sofifa","kind":"alternative","access":"public","period":null,"universe":"approximately 12,000 professional soccer players","quote":"Utilizing data from about 12,000 players from Sofifa, the Boruta algorithm streamlined feature selection."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"The Gradient Boosting Decision Tree (GBDT) model achieved the highest predictive performance with an R‑squared of 0.901 and the lowest RMSE of 3,221,632.175 on the test set.","kind":"predictive-accuracy","quote":"Notably, within the test set, the GBDT model sustained its advantage, achieving the highest R-squared value of 0.901 and the lowest RMSE of 3,221,632.175, indicative of its robustness in predicting player market values."},{"finding":"Across six ensemble algorithms, GBDT outperformed CatBoost, LightGBM, Random Forest, XGBoost, and AdaBoost in both R‑squared and RMSE metrics.","kind":"performance","quote":"Within the ensemble of six learning algorithms evaluated, the Gradient Boosting Decision Tree (GBDT) model demonstrated superior performance, achieving the highest R‑Squared value of 0.889... the"},{"finding":"Feature selection reduced the original 29 attributes to 22 key variables, including acceleration, heading accuracy, defensive awareness, and others.","kind":"statistical","quote":"Utilization of the Boruta algorithm, implemented through the BorutaShap package in Python, allowed for a reduction in the feature set to 22, as indicated by the green bars in Figure 3. Acceleration, Heading accuracy, Defensive awareness, Vision, Volleys, Sprint speed, Long passing, Positioning, Standing tackle, Dribbli"}],"reported_performance":null,"out_of_sample":{"method":"unclear","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":20,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T18:19:56Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}