# A Comparison of Machine Learning Algorithms for National Basketball Association (NBA) Most Valuable Player (MVP) Vote Share Prediction

### Zhicheng Cheng

#### 2023

#### Abstract

In an era of burgeoning sports betting, the quest to predict the Most Valuable Player (MVP) in a regular National Basketball Association (NBA) season has become a novel way for people to be involved in the world’s most popular basketball league. This paper adopts various Machine Learning Regression Models to help predict the MVP win share of an arbitrary player in an arbitrary NBA season. More specifically, every single NBA player’s statistics and MVP win share in the past 40 years are collected, preprocessed, and used to train and test the machine learning models. After comparing each model’s R-squared value and MAPE, it is concluded that the Extreme Gradient Boosting Regression Model is the best model in predicting the MVP win share of an arbitrary player in an arbitrary season, with a R-squared value of 0.6399 and a MAPE of 22.90%. This means that 63.99% of the variation in the dependent variable (i.e., the actual MVP win share) can be explained by the independent variables (the statistics), and that the prediction of the dependent variable (i.e., the actual MVP win share) is only off by 22.90%.

Download#### Paper Citation

#### in Harvard Style

Cheng Z. (2023). **A Comparison of Machine Learning Algorithms for National Basketball Association (NBA) Most Valuable Player (MVP) Vote Share Prediction**. In *Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML*; ISBN 978-989-758-705-4, SciTePress, pages 262-267. DOI: 10.5220/0012800800003885

#### in Bibtex Style

@conference{daml23,

author={Zhicheng Cheng},

title={A Comparison of Machine Learning Algorithms for National Basketball Association (NBA) Most Valuable Player (MVP) Vote Share Prediction},

booktitle={Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML},

year={2023},

pages={262-267},

publisher={SciTePress},

organization={INSTICC},

doi={10.5220/0012800800003885},

isbn={978-989-758-705-4},

}

#### in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Data Analysis and Machine Learning - Volume 1: DAML

TI - A Comparison of Machine Learning Algorithms for National Basketball Association (NBA) Most Valuable Player (MVP) Vote Share Prediction

SN - 978-989-758-705-4

AU - Cheng Z.

PY - 2023

SP - 262

EP - 267

DO - 10.5220/0012800800003885

PB - SciTePress