Paired Comparison Models for Ranking Major League Baseball Teams

碩士 === 中華大學 === 應用統計學系碩士班 === 100 === This thesis studies the within-pair order effect of paired comparison tied models. The object of study is the U.S. Major League Baseball. The main application of paired comparison model is to identify the winning probability and the ranking for all the teams. Th...

Full description

Bibliographic Details
Main Authors: Chan, Lan-Ti, 詹嵐迪
Other Authors: Chi Lo
Format: Others
Language:zh-TW
Published: 2012
Online Access:http://ndltd.ncl.edu.tw/handle/64524339156365530296
id ndltd-TW-100CHPI5506016
record_format oai_dc
spelling ndltd-TW-100CHPI55060162015-10-13T21:17:10Z http://ndltd.ncl.edu.tw/handle/64524339156365530296 Paired Comparison Models for Ranking Major League Baseball Teams 配對比較模式的研究-以美國職棒大聯盟的排序為例 Chan, Lan-Ti 詹嵐迪 碩士 中華大學 應用統計學系碩士班 100 This thesis studies the within-pair order effect of paired comparison tied models. The object of study is the U.S. Major League Baseball. The main application of paired comparison model is to identify the winning probability and the ranking for all the teams. The article introduces the origin of all models and the relationship between all the models. Then the parameters of the model are estimated and associated hypotheses are tested. Finally, the information from collected data is analyzed by the Bradley-Terry model, Davidson, model and multiplicative order effect Davidson model to incorporate ties and conclusions are made. In the Bradley-Terry model, the “true” ratings of 14 teams in American League are not all equal. The result is Yankees ranked the first and the Angels ranked the second and the last one is the Orioles. There is no significant difference in the ratings between Yankees and Angels. The “true” ratings of 16 teams in National League are not all equal. The result is Cardinals ranked the first and the Phillies ranked the second and the last one is the Nationals. There is no significant difference in the ratings between Cardinals and Phillies. The “true” ratings of 30 teams in American and National League are not all equal. The result is Yankees ranked the first and the Angels ranked the second and the last one is the Nationals. There is no significant difference in the ratings between Yankees and Angels. American League ranking does not change in the Davidson model to consider a tie. The first rank is Phillies and the second rank is Dodgers in National League. The ranking of the American League and National League have no change. In the Davidson model incorporating the tie and the multiplicative order effect, American League ranking does not change. The first rank is Phillies and the second rank is Dodgers in National League. The ranking of the American League and National League have no change. Yankees ranked the first and the Angels ranked the second and the last one is the Nationals. Chi Lo 羅琪 2012 學位論文 ; thesis 154 zh-TW
collection NDLTD
language zh-TW
format Others
sources NDLTD
description 碩士 === 中華大學 === 應用統計學系碩士班 === 100 === This thesis studies the within-pair order effect of paired comparison tied models. The object of study is the U.S. Major League Baseball. The main application of paired comparison model is to identify the winning probability and the ranking for all the teams. The article introduces the origin of all models and the relationship between all the models. Then the parameters of the model are estimated and associated hypotheses are tested. Finally, the information from collected data is analyzed by the Bradley-Terry model, Davidson, model and multiplicative order effect Davidson model to incorporate ties and conclusions are made. In the Bradley-Terry model, the “true” ratings of 14 teams in American League are not all equal. The result is Yankees ranked the first and the Angels ranked the second and the last one is the Orioles. There is no significant difference in the ratings between Yankees and Angels. The “true” ratings of 16 teams in National League are not all equal. The result is Cardinals ranked the first and the Phillies ranked the second and the last one is the Nationals. There is no significant difference in the ratings between Cardinals and Phillies. The “true” ratings of 30 teams in American and National League are not all equal. The result is Yankees ranked the first and the Angels ranked the second and the last one is the Nationals. There is no significant difference in the ratings between Yankees and Angels. American League ranking does not change in the Davidson model to consider a tie. The first rank is Phillies and the second rank is Dodgers in National League. The ranking of the American League and National League have no change. In the Davidson model incorporating the tie and the multiplicative order effect, American League ranking does not change. The first rank is Phillies and the second rank is Dodgers in National League. The ranking of the American League and National League have no change. Yankees ranked the first and the Angels ranked the second and the last one is the Nationals.
author2 Chi Lo
author_facet Chi Lo
Chan, Lan-Ti
詹嵐迪
author Chan, Lan-Ti
詹嵐迪
spellingShingle Chan, Lan-Ti
詹嵐迪
Paired Comparison Models for Ranking Major League Baseball Teams
author_sort Chan, Lan-Ti
title Paired Comparison Models for Ranking Major League Baseball Teams
title_short Paired Comparison Models for Ranking Major League Baseball Teams
title_full Paired Comparison Models for Ranking Major League Baseball Teams
title_fullStr Paired Comparison Models for Ranking Major League Baseball Teams
title_full_unstemmed Paired Comparison Models for Ranking Major League Baseball Teams
title_sort paired comparison models for ranking major league baseball teams
publishDate 2012
url http://ndltd.ncl.edu.tw/handle/64524339156365530296
work_keys_str_mv AT chanlanti pairedcomparisonmodelsforrankingmajorleaguebaseballteams
AT zhānlándí pairedcomparisonmodelsforrankingmajorleaguebaseballteams
AT chanlanti pèiduìbǐjiàomóshìdeyánjiūyǐměiguózhíbàngdàliánméngdepáixùwèilì
AT zhānlándí pèiduìbǐjiàomóshìdeyánjiūyǐměiguózhíbàngdàliánméngdepáixùwèilì
_version_ 1718058798265925632