A Logistic Regression/Markov Chain Model for American College Football

Kvam and Sokol developed a successful logistic regression/Markov chain (LRMC) model for ranking college basketball teams part of Division I of the National Colligate Athletic Association (NCAA). In their 2006 publication, they illustrated that the LRMC model is one of the most successful ranking sys...

Full description

Saved in:
Bibliographic Details
Main Authors: Kolbush J., Sokol J.
Format: Article
Language:English
Published: Sciendo 2017-12-01
Series:International Journal of Computer Science in Sport
Subjects:
Online Access:https://doi.org/10.1515/ijcss-2017-0014
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850060942376697856
author Kolbush J.
Sokol J.
author_facet Kolbush J.
Sokol J.
author_sort Kolbush J.
collection DOAJ
description Kvam and Sokol developed a successful logistic regression/Markov chain (LRMC) model for ranking college basketball teams part of Division I of the National Colligate Athletic Association (NCAA). In their 2006 publication, they illustrated that the LRMC model is one of the most successful ranking systems in predicting the outcome of the NCAA Division I Basketball Tournament. However, it cannot directly be extended to college football because of the lack of home-and-home matchups that LRMC exploits in performing its Logistic Regression. We present a common-opponents-based approach that allows us to perform a Logistic Regression and thus create a football LRMC (F-LRMC) model. This approach compares the margin of victory of home teams to their winning percentage in games played against common-opponents with the away team. Computational results show that F-LRMC is among the best of the many ranking systems tracked by Massey's College Football Ranking Composite.
format Article
id doaj-art-072a128c3f274e3eb3d97e4356b01ecf
institution DOAJ
issn 1684-4769
language English
publishDate 2017-12-01
publisher Sciendo
record_format Article
series International Journal of Computer Science in Sport
spelling doaj-art-072a128c3f274e3eb3d97e4356b01ecf2025-08-20T02:50:26ZengSciendoInternational Journal of Computer Science in Sport1684-47692017-12-0116318519610.1515/ijcss-2017-0014A Logistic Regression/Markov Chain Model for American College FootballKolbush J.0Sokol J.1H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of TechnologyH. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of TechnologyKvam and Sokol developed a successful logistic regression/Markov chain (LRMC) model for ranking college basketball teams part of Division I of the National Colligate Athletic Association (NCAA). In their 2006 publication, they illustrated that the LRMC model is one of the most successful ranking systems in predicting the outcome of the NCAA Division I Basketball Tournament. However, it cannot directly be extended to college football because of the lack of home-and-home matchups that LRMC exploits in performing its Logistic Regression. We present a common-opponents-based approach that allows us to perform a Logistic Regression and thus create a football LRMC (F-LRMC) model. This approach compares the margin of victory of home teams to their winning percentage in games played against common-opponents with the away team. Computational results show that F-LRMC is among the best of the many ranking systems tracked by Massey's College Football Ranking Composite.https://doi.org/10.1515/ijcss-2017-0014logistic regresionmarkov chainamerican college footballcommon gamemargin of victory
spellingShingle Kolbush J.
Sokol J.
A Logistic Regression/Markov Chain Model for American College Football
International Journal of Computer Science in Sport
logistic regresion
markov chain
american college football
common game
margin of victory
title A Logistic Regression/Markov Chain Model for American College Football
title_full A Logistic Regression/Markov Chain Model for American College Football
title_fullStr A Logistic Regression/Markov Chain Model for American College Football
title_full_unstemmed A Logistic Regression/Markov Chain Model for American College Football
title_short A Logistic Regression/Markov Chain Model for American College Football
title_sort logistic regression markov chain model for american college football
topic logistic regresion
markov chain
american college football
common game
margin of victory
url https://doi.org/10.1515/ijcss-2017-0014
work_keys_str_mv AT kolbushj alogisticregressionmarkovchainmodelforamericancollegefootball
AT sokolj alogisticregressionmarkovchainmodelforamericancollegefootball
AT kolbushj logisticregressionmarkovchainmodelforamericancollegefootball
AT sokolj logisticregressionmarkovchainmodelforamericancollegefootball