A Statistical Investigation into Factors Affecting Results of One Day International Cricket Matches

The effect of playing ‟home” or ‟away” and many other factors, such as batting first or second, winning or losing the toss, have been hypothesised as influencing the outcome of major cricket matches. Anecdotally, it has often been noted that Subcontinental sides (India, Pakistan, Sri Lanka and Bangladesh) tend to perform much better on the Subcontinent than away from it, whilst England do better in Australia during cooler, damper Australian Summers than during hotter, drier ones. In this paper, focusing on results of men‘s One Day International (ODI) matches involving England, we investigate the extent to which a number of factors - including playing home or away (or the continent of the venue), batting or fielding first, winning or losing the toss, the weather conditions during the game, the condition of the pitch, and the strength of each team‘s top batting and bowling resources - influence the outcome of matches. By employing a variety of Statistical techniques, we find that the continent of the venue does appear to be a major factor affecting the result, but winning the toss does not. We then use the factors identified as significant in an attempt to build a Binary Logistic Regression Model that will estimate the probability of England winning at various stages of a game. Finally, we use this model to predict the results of some England ODI games not used in training the model.