Home > Research > Publications & Outputs > A state-space perspective on modelling and infe...

Links

Text available via DOI:

View graph of relations

A state-space perspective on modelling and inference for online skill rating

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

A state-space perspective on modelling and inference for online skill rating. / Duffield, S.; Power, S.; Rimella, L.
In: Journal of the Royal Statistical Society: Series C (Applied Statistics), Vol. 73, No. 5, 30.11.2024, p. 1262-1282.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Duffield, S, Power, S & Rimella, L 2024, 'A state-space perspective on modelling and inference for online skill rating', Journal of the Royal Statistical Society: Series C (Applied Statistics), vol. 73, no. 5, pp. 1262-1282. https://doi.org/10.1093/jrsssc/qlae035

APA

Duffield, S., Power, S., & Rimella, L. (2024). A state-space perspective on modelling and inference for online skill rating. Journal of the Royal Statistical Society: Series C (Applied Statistics), 73(5), 1262-1282. https://doi.org/10.1093/jrsssc/qlae035

Vancouver

Duffield S, Power S, Rimella L. A state-space perspective on modelling and inference for online skill rating. Journal of the Royal Statistical Society: Series C (Applied Statistics). 2024 Nov 30;73(5):1262-1282. Epub 2024 Aug 16. doi: 10.1093/jrsssc/qlae035

Author

Duffield, S. ; Power, S. ; Rimella, L. / A state-space perspective on modelling and inference for online skill rating. In: Journal of the Royal Statistical Society: Series C (Applied Statistics). 2024 ; Vol. 73, No. 5. pp. 1262-1282.

Bibtex

@article{9737a8a81f7449e1a6aa74c6ce3e4f0a,
title = "A state-space perspective on modelling and inference for online skill rating",
abstract = "We summarize popular methods used for skill rating in competitive sports, along with their inferential paradigms and introduce new approaches based on sequential Monte Carlo and discrete hidden Markov models. We advocate for a state-space model perspective, wherein players{\textquoteright} skills are represented as time-varying, and match results serve as observed quantities. We explore the steps to construct the model and the three stages of inference: filtering, smoothing, and parameter estimation. We examine the challenges of scaling up to numerous players and matches, highlighting the main approximations and reductions which facilitate statistical and computational efficiency. We additionally compare approaches in a realistic experimental pipeline that can be easily reproduced and extended with our open-source Python package, abile. {\textcopyright} The Royal Statistical Society 2024.",
keywords = "approximate inference, Bayesian inference, competitive sports, state-space models",
author = "S. Duffield and S. Power and L. Rimella",
note = "Export Date: 28 November 2024 Correspondence Address: Power, S.; School of Mathematics, Fry Building, Woodland Road, United Kingdom; email: sam.power@bristol.ac.uk Funding details: Engineering and Physical Sciences Research Council, EPSRC, EP/R018561/1 Funding text 1: S. Power and L. Rimella were supported by EPSRC grant EP/R018561/1 (Bayes4Health).",
year = "2024",
month = nov,
day = "30",
doi = "10.1093/jrsssc/qlae035",
language = "English",
volume = "73",
pages = "1262--1282",
journal = "Journal of the Royal Statistical Society: Series C (Applied Statistics)",
issn = "0035-9254",
publisher = "Wiley-Blackwell",
number = "5",

}

RIS

TY - JOUR

T1 - A state-space perspective on modelling and inference for online skill rating

AU - Duffield, S.

AU - Power, S.

AU - Rimella, L.

N1 - Export Date: 28 November 2024 Correspondence Address: Power, S.; School of Mathematics, Fry Building, Woodland Road, United Kingdom; email: sam.power@bristol.ac.uk Funding details: Engineering and Physical Sciences Research Council, EPSRC, EP/R018561/1 Funding text 1: S. Power and L. Rimella were supported by EPSRC grant EP/R018561/1 (Bayes4Health).

PY - 2024/11/30

Y1 - 2024/11/30

N2 - We summarize popular methods used for skill rating in competitive sports, along with their inferential paradigms and introduce new approaches based on sequential Monte Carlo and discrete hidden Markov models. We advocate for a state-space model perspective, wherein players’ skills are represented as time-varying, and match results serve as observed quantities. We explore the steps to construct the model and the three stages of inference: filtering, smoothing, and parameter estimation. We examine the challenges of scaling up to numerous players and matches, highlighting the main approximations and reductions which facilitate statistical and computational efficiency. We additionally compare approaches in a realistic experimental pipeline that can be easily reproduced and extended with our open-source Python package, abile. © The Royal Statistical Society 2024.

AB - We summarize popular methods used for skill rating in competitive sports, along with their inferential paradigms and introduce new approaches based on sequential Monte Carlo and discrete hidden Markov models. We advocate for a state-space model perspective, wherein players’ skills are represented as time-varying, and match results serve as observed quantities. We explore the steps to construct the model and the three stages of inference: filtering, smoothing, and parameter estimation. We examine the challenges of scaling up to numerous players and matches, highlighting the main approximations and reductions which facilitate statistical and computational efficiency. We additionally compare approaches in a realistic experimental pipeline that can be easily reproduced and extended with our open-source Python package, abile. © The Royal Statistical Society 2024.

KW - approximate inference

KW - Bayesian inference

KW - competitive sports

KW - state-space models

U2 - 10.1093/jrsssc/qlae035

DO - 10.1093/jrsssc/qlae035

M3 - Journal article

VL - 73

SP - 1262

EP - 1282

JO - Journal of the Royal Statistical Society: Series C (Applied Statistics)

JF - Journal of the Royal Statistical Society: Series C (Applied Statistics)

SN - 0035-9254

IS - 5

ER -