Answers are grounded in this paper and cite their sources.
Negative binomial and Poisson distributions are fitted to data on scores in Association Football for the seasons 1946–47 to 1983–84. There are strong grounds for preferring the negative binomial up to 1970; thereafter the Poisson seems adequate. Simplification is achieved by fitting the negative binomial with a common parameter. The analyses are set in the context of previous applications and interpretations in the area. Different models giving rise to the negative binomial or Poisson are investigated and some support found for models not previously advanced in this context. Notwith-standing the success of such exercises some scepticism is expressed about the interpretations placed on previous analyses.
No takes yet. Share an insight, caveat, or question.
Baxter et al. (1988) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: