TY - JOUR
T1 - The predictive power of popular sports ranking methods in the NFL, NBA, and NHL
AU - Dabadghao, S.S.
AU - Vaziri, B.
N1 - Publisher Copyright:
© 2021, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.
Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.
PY - 2022/7
Y1 - 2022/7
N2 - Ideally, a ranking method for a sports tournament will be not only fair and comprehensive, but it will also possess strong predictive power. In a recent article, the (1 , α) method was proposed as being fair and comprehensive in comparison with five other popular sports ranking methods. In this study, we will compare the predictive power of the (1 , α) method against five popular sports ranking methods (Win-loss, Massey, Colley, Markov, and Elo) for the NFL, NBA, and NHL in the last two decades. We use regular season results to obtain the ranking vector, and then evaluate said ranking vector on its ability to predict playoff results using both a frequency count and McNemar’s test of statistical significance.
AB - Ideally, a ranking method for a sports tournament will be not only fair and comprehensive, but it will also possess strong predictive power. In a recent article, the (1 , α) method was proposed as being fair and comprehensive in comparison with five other popular sports ranking methods. In this study, we will compare the predictive power of the (1 , α) method against five popular sports ranking methods (Win-loss, Massey, Colley, Markov, and Elo) for the NFL, NBA, and NHL in the last two decades. We use regular season results to obtain the ranking vector, and then evaluate said ranking vector on its ability to predict playoff results using both a frequency count and McNemar’s test of statistical significance.
KW - Markov
KW - Predictive power
KW - Ranking
KW - Sports
UR - https://www.scopus.com/pages/publications/85102785971
U2 - 10.1007/s12351-021-00630-9
DO - 10.1007/s12351-021-00630-9
M3 - Article
AN - SCOPUS:85102785971
SN - 1109-2858
VL - 22
SP - 2767
EP - 2783
JO - Operational Research
JF - Operational Research
IS - 3
ER -