Predicting match outcomes in the 2006 football world cup considering time-effect weights
Resumo
In this paper we propose a bayesian subjective statistical methodology for predicting match outcomes of 2006 World Cup
Association Football. The prior information considered by the model is the opinion of experts and the FIFA scores announced
previously to the start of the competition. As matches occur, the observed results are included in themodel with different weights, which are inversely proportional to the elapsed time of the observation. In addition to the weights given to previous results, a weight is associated with the opinions of experts. This approach allows for the model calibration, directed to increase
the predictive capability, through the appropriate choice of the weights associated with the experts’ opinions and with the previous matches. The win, draw and loss probabilities for each match were obtained exactly and, by means of a stochastic simulation, we estimated the probabilities of classification in the group stage, to reach each round of the knockout stage and
winning the tournament.
Association Football. The prior information considered by the model is the opinion of experts and the FIFA scores announced
previously to the start of the competition. As matches occur, the observed results are included in themodel with different weights, which are inversely proportional to the elapsed time of the observation. In addition to the weights given to previous results, a weight is associated with the opinions of experts. This approach allows for the model calibration, directed to increase
the predictive capability, through the appropriate choice of the weights associated with the experts’ opinions and with the previous matches. The win, draw and loss probabilities for each match were obtained exactly and, by means of a stochastic simulation, we estimated the probabilities of classification in the group stage, to reach each round of the knockout stage and
winning the tournament.
Palavras-chave
Bayesian Inference, Simulation, Sport forecasting, World Cup Association Football.
Apontamentos
- Não há apontamentos.
Este trabalho está licenciado sob uma Licença Creative Commons Attribution 3.0 .
- Auxílio à publicação da Fundação de Amparo à Pesquisa do Estado de Minas Gerais – FAPEMIG, edital 05/2014.
- Processo: APL-00063-14