Pitcher Performance Prediction Major League Baseball (MLB) by Temporal Fusion Transformer
Source abstract
Abstract
Predicting player performance in sports is a critical challenge with significant implications for team success, fan engagement, and financial outcomes. Although, in Major League Baseball (MLB), statistical methodologies such as sabermetrics have been widely used, the dynamic nature of sports makes accurate performance prediction a difficult task. Enhanced forecasts can provide immense value to team managers by aiding strategic player contract and acquisition decisions. This study addresses this challenge by employing the temporal fusion transformer (TFT), an advanced and cutting-edge deep learning model for complex data, to predict pitchers’ earned run average (ERA), a key metric in baseball performance analysis. The performance of the TFT model is evaluated against recurrent neural network-based approaches and existing projection systems. In experimental results, the TFT based model consistently outperformed its counterparts, demonstrating superior accuracy in pitcher performance prediction. By leveraging the advanced capabilities of TFT, this study contributes to more precise player evaluations and improves strategic planning in baseball.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Sports Analytics and Performance · Music and Audio Processing
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (53)
Show source-provided references
1. Baseball predictions and strategies using explainable AI
2. High-school baseball pitcher’s ERA (Earned Run Average) prediction using multi-variable linear regression analysis method
3. Sabermetric analysis: wins-above-replacement
4. Moorefield J. The Oakland Athletics use of sabermetrics and the rise of big data analytics in business [master’s thesis]. Chattanooga, TN, USA: University of Tennessee at Chattanooga; 2021.
5. Gate-variants of gated recurrent unit (GRU) neural networks
6. Graves · Long short-term memory. Supervised sequence labelling with recurrent neural networks · 2012 · 3710.1007/978-3-642-24797-2_4 ↗
7. Bi-LSTM-based two-stream network for machine remaining useful life prediction
8. Deep learning with long short-term memory networks for financial market predictions10.1016/j.ejor.2017.11.054 ↗
9. Temporal fusion transformers for interpretable multi-horizon time series forecasting10.1016/j.ijforecast.2021.03.012 ↗
10. Designing a prediction model for athlete’s sports performance using neural network10.1007/s00500-023-09091-y ↗
11. Selection procedures in sports: improving predictions of athletes’ future performance10.1080/17461391.2018.1480662 ↗
12. A compound framework for sports prediction: the case study of football10.1016/j.knosys.2008.03.016 ↗
13. Sports performance prediction based on chaos theory and machine learning10.1155/2022/3916383 ↗
14. Prediction model and data simulation of sports performance based on the artificial intelligence algorithm10.1155/2022/7238789 ↗
15. Sports data mining for cricket match prediction
16. Prediction of marathon performance using artificial intelligence10.1055/a-1993-2371 ↗
17. Use of machine learning and deep learning to predict the outcomes of major league baseball matches10.3390/app11104499 ↗
18. Cui AY. Forecasting outcomes of major league baseball games using machine learning [master’s thesis]. Philadelphia, PA, USA: University of Pennsylvania; 2020.
19. Exploring and selecting features to predict the next outcomes of MLB games10.3390/e24020288 ↗
20. Performance prediction in major league baseball by long short-term memory networks10.1007/s41060-022-00313-4 ↗
21. Predicting home run production in Major League Baseball using a Bayesian semiparametric model10.1080/00031305.2017.1401959 ↗
22. The prediction of batting averages in major league baseball10.3390/stats3020008 ↗
23. Machine learning applications in baseball: a systematic literature review10.1080/08839514.2018.1442991 ↗
24. Statcast dashboard: exploration of spatiotemporal baseball data10.1109/MCG.2016.101 ↗
25. Statcast and the baseball trajectory calculator10.1119/1.4976652 ↗
26. Sports knowledge management and data mining10.1002/aris.2010.1440440110 ↗
27. A review of data mining techniques for result prediction in sports
28. Sport-specific practice and the development of expert decision-making in team ball sports10.1080/10413200305400 ↗
29. Performance analysis for decision making in team sports
30. Enhancing decision-making during sports performance: current understanding and future directions10.53841/bpssepr.2016.12.1.54 ↗
31. Mittal H, Rikhari D, Kumar J, Singh AK. A study on machine learning approaches for player performance and match results prediction. arXiv:210810125. 2021.
32. Prediction of athletes performance using neural networks: an application in cricket team selection10.1016/j.eswa.2008.06.088 ↗
33. A review of recurrent neural networks: LSTM cells and network architectures10.1162/neco_a_01199 ↗
34. Learning to forget: continual prediction with LSTM10.1162/089976600300015015 ↗
35. Depression prediction based on BiAttention-GRU10.1007/s12652-021-03497-y ↗
36. Air pollution forecasting using a deep learning model based on 1D convnets and bidirectional GRU10.1109/ACCESS.2019.2921578 ↗
37. Attention-based bidirectional LSTM-CNN model for remaining useful life estimation
38. Remaining useful life estimation using a bidirectional recurrent neural network based autoencoder scheme10.1016/j.ymssp.2019.05.005 ↗
39. Comparing bitcoin’s prediction model using GRU, RNN, and LSTM by hyperparameter optimization grid search and random search
40. Intelligent traffic flow prediction using optimized GRU model10.1109/ACCESS.2021.3097141 ↗
41. Interpretable wind speed prediction with multivariate time series and temporal fusion transformers10.1016/j.energy.2022.123990 ↗
42. A temporal fusion transformer for short-term freeway traffic speed multistep prediction10.1016/j.neucom.2022.05.083 ↗
43. Stock price prediction based on temporal fusion transformer
44. MultiMix TFT: a multi-task mixed-frequency framework with temporal fusion transformers
45. Lundberg S. A unified approach to interpreting model predictions. arXiv:170507874. 2017.
46. Christoph · Interpretable machine learning: a guide for making black box models explainable · 202010.58248/PN633 ↗
47. Lundberg SM, Erion GG, Lee SI. Consistent individualized feature attribution for tree ensembles. arXiv:180203888. 2018.
48. Recent advances in predictive learning analytics: a decade systematic review (2012–2022)10.1007/s10639-022-11536-0 ↗
49. Kamarthi H, Rodríguez A, Prakash BA. Back2Future: leveraging backfill dynamics for improving real-time predictions in future. arXiv:210604420. 2021.
50. Short-term memory for serial order: a recurrent neural network model10.1037/0033-295X.113.2.201 ↗
51. Variš D, Bojar O. Sequence length is a domain: length-based overfitting in transformer models. arXiv:210907276. 2021.10.18653/v1/2021.emnlp-main.650 ↗
52. Informer: beyond efficient transformer for long sequence time-series forecasting
53. Zhang X, Ghosh S. PaEBack: pareto-efficient backsubsampling for time series data. arXiv:221015780. 2022.