Modelling the evolution of distributions

an application to major league baseball

Research output: Contribution to journalArticle

5 Citations (Scopus)
94 Downloads (Pure)

Abstract

We develop Bayesian techniques for modelling the evolution of entire distributions over time and apply them to the distribution of team performance in Major League baseball for the period 1901-2000. Such models offer insight into many key issues (e.g. competitive balance) in a way that regression-based models cannot. The models involve discretizing the distribution and then modelling the evolution of the bins over time through transition probability matrices. We allow for these matrices to vary over time and across teams. We find that, with one exception, the transition probability matrices (and, hence, competitive balance) have been remarkably constant across time and over teams. The one exception is the Yankees, who have outperformed all other teams.
Original languageEnglish
Pages (from-to)639-656
Number of pages18
JournalJournal of the Royal Statistical Society: Series A
Volume167
Issue number4
DOIs
Publication statusPublished - Nov 2004

Fingerprint

Transition Probability Matrix
Modeling
Exception
Time Constant
Regression
Entire
Vary
Model
regression
Major League Baseball
time
performance
Competitive balance
Transition probability
Team performance

Keywords

  • bayesian
  • Gibbs sampler
  • ordered probit
  • sports statistics

Cite this

@article{05048d5a2c16407e8b082493c7528558,
title = "Modelling the evolution of distributions: an application to major league baseball",
abstract = "We develop Bayesian techniques for modelling the evolution of entire distributions over time and apply them to the distribution of team performance in Major League baseball for the period 1901-2000. Such models offer insight into many key issues (e.g. competitive balance) in a way that regression-based models cannot. The models involve discretizing the distribution and then modelling the evolution of the bins over time through transition probability matrices. We allow for these matrices to vary over time and across teams. We find that, with one exception, the transition probability matrices (and, hence, competitive balance) have been remarkably constant across time and over teams. The one exception is the Yankees, who have outperformed all other teams.",
keywords = "bayesian, Gibbs sampler, ordered probit, sports statistics",
author = "Gary Koop",
note = "Working paper version",
year = "2004",
month = "11",
doi = "10.1111/j.1467-985X.2004.A473.x",
language = "English",
volume = "167",
pages = "639--656",
journal = "Journal of the Royal Statistical Society: Series A",
issn = "0964-1998",
number = "4",

}

Modelling the evolution of distributions : an application to major league baseball. / Koop, Gary.

In: Journal of the Royal Statistical Society: Series A , Vol. 167, No. 4, 11.2004, p. 639-656.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Modelling the evolution of distributions

T2 - an application to major league baseball

AU - Koop, Gary

N1 - Working paper version

PY - 2004/11

Y1 - 2004/11

N2 - We develop Bayesian techniques for modelling the evolution of entire distributions over time and apply them to the distribution of team performance in Major League baseball for the period 1901-2000. Such models offer insight into many key issues (e.g. competitive balance) in a way that regression-based models cannot. The models involve discretizing the distribution and then modelling the evolution of the bins over time through transition probability matrices. We allow for these matrices to vary over time and across teams. We find that, with one exception, the transition probability matrices (and, hence, competitive balance) have been remarkably constant across time and over teams. The one exception is the Yankees, who have outperformed all other teams.

AB - We develop Bayesian techniques for modelling the evolution of entire distributions over time and apply them to the distribution of team performance in Major League baseball for the period 1901-2000. Such models offer insight into many key issues (e.g. competitive balance) in a way that regression-based models cannot. The models involve discretizing the distribution and then modelling the evolution of the bins over time through transition probability matrices. We allow for these matrices to vary over time and across teams. We find that, with one exception, the transition probability matrices (and, hence, competitive balance) have been remarkably constant across time and over teams. The one exception is the Yankees, who have outperformed all other teams.

KW - bayesian

KW - Gibbs sampler

KW - ordered probit

KW - sports statistics

U2 - 10.1111/j.1467-985X.2004.A473.x

DO - 10.1111/j.1467-985X.2004.A473.x

M3 - Article

VL - 167

SP - 639

EP - 656

JO - Journal of the Royal Statistical Society: Series A

JF - Journal of the Royal Statistical Society: Series A

SN - 0964-1998

IS - 4

ER -