CoChat

Predictions of european basketball match results with machine learning algorithms

Open access
Lampis, Tzai, Ioannis, Ntzoufras, Vasilios, Vassalos, Stavrianna, Dimitriou · 2023 · Journal of Sports Analytics

The DOI matches its Crossref record; 31 of the 31 references it cites were found in a registry or an index; 1 of the 9 sources its claims cite were checked for support.

Abstract

The goal of this paper is to build and compare methods for the prediction of the final outcomes of basketball games. In this study, we analyzed data from four different European tournaments: Euroleague, Eurocup, Greek Basket League and Spanish Liga ACB. The data-set consists of information collected from box scores of 5214 games for the period of 2013-2018. The predictions obtained by our implemented methods and models were compared with a “vanilla” model using only the team-name information of each game. In our analysis, we have included new performance indicators constructed by using historical statistics, key performance indicators and measurements from three rating systems (Elo, PageRank, pi-rating). For these three rating systems and every tournament under consideration, we tune the rating system parameters using specific training data-sets. These new game features are improving our predictions efficiently and can be easily obtained in any basketball league. Our predictions were obtained by implementing three different statistics and machine learning algorithms: logistic regression, random forest, and extreme gradient boosting trees. Moreover, we report predictions based on the combination of these algorithms (ensemble learning). We evaluate our predictions using three predictive measures: Brier Score, accuracy and F 1 -score. In addition, we evaluate the performance of our algorithms with three different prediction scenarios (full-season, mid-season, and play-offs predictive evaluation). For the mid-season and the play-offs scenarios, we further explore whether incorporating additional results from previous seasons in the learning data-set enhances the predictive performance of the implemented models and algorithms. Concerning the results, there is no clear winner between the machine learning algorithms since they provide identical predictions with small differences. However, models with predictors suggested in this paper out-perform the “vanilla” model by 3-5% in terms of accuracy. Another conclusion from our results for the play-offs scenarios is that it is not necessary to embed outcomes from previous seasons in our training data-set. Using data from the current season, most of the time, leads to efficient, accurate parameter learning and well-behaved prediction models. Moreover, the Greek league is the least balanced tournament in terms of competitiveness since all our models achieve high predictive accuracy (78%, on the best-performing model). The second less balanced league is the Spanish one with accuracy reaching 72% while for the two European tournaments the prediction accuracy is considerably lower (about 69%). Finally, we present the most important features by counting the percentage of appearance in every machine learning algorithm for every one of the three analyses. From this analysis, we may conclude that the best predictors are the rating systems (pi-rating, PageRank, and ELO) and the current form performance indicators (e.g., the two most frequent ones are the game score of Hollinger and the floor impact counter).

Overview

CoChat Summary

TL;DR

The paper builds and compares machine-learning models to predict winners of European basketball games across four tournaments, using box-score, rating-system, current-form, and tournament features. It finds that richer feature models improve on a team-name-only baseline by about 3–5% in accuracy, but no single algorithm clearly dominates.

Key findings

  • The Full Information Model outperformed the Baseline Vanilla Model by 3–5% in accuracy when using all previous three seasons to predict the last available season; improvements were also similar for mid-season and play-off scenarios, with some exceptions.
  • No clear winner emerged among logistic regression, random forest, XGBoost, and ensemble learning: “they provide identical predictions with small differences” when large training data sets are used.
  • The Greek league was the most predictable/least balanced tournament, with best-model accuracy of 78%; the Spanish Liga ACB reached 72%, while the two European tournaments were lower at about 69%.
  • The most consistent important predictors were rating systems (pi-rating, PageRank, and ELO) and current-form performance indicators; ELO appeared as important in about 60% of examined prediction scenarios.

Methods

  • Data: 5,214 games from four European tournaments—Euroleague, Eurocup, Greek Basket League, and Spanish Liga ACB—over seasons 2013–2018; target was a binary home-team win indicator.
  • Features: 110 final features across historical information, rating systems (ELO, PageRank, pi-rating), current form over the last 10 games, and tournament-specific characteristics; arithmetic features were min-max normalized to [0,1] within each tournament.
  • Models and evaluation: logistic regression with regularization, random forest, XGBoost, and ensemble averaging; tuning used randomized search with 10-fold cross-validation and Brier score, with final evaluation by Brier score, accuracy, and F1 across full-season, mid-season, and play-off scenarios.

Limitations

  • The paper states that team rosters frequently change between seasons, and new teams lack historical information; for a team’s first game, all features were set to zero, affecting Euroleague 2.16%, Eurocup 6.6%, Greek League 1.82%, and Liga ACB 0.32% of cases.
  • The paper notes that the large number of features creates a multiplicity problem and possible false-positive important features, and that a weighting scheme giving more importance to recent matches would have been more appropriate but was not implemented.
  • The paper states that the play-off test set is small, so evaluation metrics can be highly variable and influenced by one game; in some cases the Baseline Vanilla Model outperformed the Full Information Model, including Liga ACB play-offs and mid-season Euroleague.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3233/jsa-220639

Cites

31 of 31 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

1 Valid · 1 Not valid · 6 Not checked · 105 Own results and interpretations

76 more claims

Create a free account to see all 113 claims and to verify the 1 reference still open.

See all claims — free account
  • It may state two separate claims as one. 18
  • The quote only partly states this claim. 1
  • Probably not a substantial claim. 14

Create a free account to see these claims.

Citations

Cites

Show all 31 references

Research works printed but not linked: 11

  1. Carlin B. P., 2005, SIAM, 149
  2. Chen T., 2016, Proceedings of the nd acm sigkdd internationalconference on knowledge discovery and data mining, 785
  3. Fan R.-E., 2008, Journal of Machine Learning Research, 9, 1871
  4. Friedman J. H., 2001, Annals of statistics, 1189
  5. Harville D. A., 1994, The AmericanStatistician, 48, 22
  6. 10.1007/s10994-018-5704-6 · DOI
  7. Li B., 1984, Biometrics, 40, 358
  8. Milanović, D., Selmanović, A. & Škegro, D., 2014, Characteristics and differences of basic types of offenses in european and american top-level basketball, in 7th International Scientific Conference on Kinesiology, pp. 400.
  9. Naismith, J., 1941, Basketball: Its Origin and Development, New York, Association Press.
  10. Shi, Z., Moorthy, S. & Zimmermann, A., 2013, Predicting ncaab match outcomes using ml techniques-some results and lessons learned, in ECML/PKDD 2013Workshop on Machine Learning and Data Mining for Sports Analytics.
  11. Van Rijsbergen, C. J., 1979, Information Retrieval, 2nd edition, Butterworths.

Cited by

See all 34

Metadata

Authors
Lampis, T., Ioannis, N., Vasilios, V., & Stavrianna, D.
Affiliations
  • Athens University of Economics and Business, Athens, Greece
  • Department of Statistics, AUEB Sports Analytics Group, Computational and Bayesian Statistics Lab, Athens University of Economics and Business, Athens, Greece
  • Athens University of Economics and Business (GR)
  • Data Strategy and Analysis, PPC S.A., Athens, Greece
Venue
Journal of Sports Analytics
Published
2023-03-31
Volume
9
Issue
2
Pages
171-190
ISSN
  • 2215-020X (ISSN-L)
  • 2215-0218
OpenAlex
W4362470549
Cited by
34
References
23
Citations by year
2026: 13 · 2025: 13 · 2024: 8
FWCI
6.01
Type
journal article
Language
en
Open access
Locations
  • Journal of Sports Analytics (publishedVersion, cc-by-nc)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Sports Analytics and Performance
  • Health Sciences › Medicine › Orthopedics and Sports Medicine › Sports Performance and Training
  • Social Sciences › Decision Sciences › Management Science and Operations Research › Forecasting Techniques and Applications
Keywords
sports prediction, Elo rating, PageRank, performance indicators, box scores, Euroleague, ensemble learning, random forest, extreme gradient boosting, logistic regression, feature importance

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Predictions of european basketball match results with machine learning algorithms. https://app.cochat.ai/papers/10.3233/jsa-220639
The paper
Lampis, T., Ioannis, N., Vasilios, V., & Stavrianna, D. (2023). Predictions of european basketball match results with machine learning algorithms. Journal of Sports Analytics. https://doi.org/10.3233/jsa-220639

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.