In the sport of cricket, recent performances, or "form", is frequently cited by selectors, coaches, players, and spectators, as a major reason for an individual's inclusion or exclusion from an upcoming match or tour. While cricket is sport with a strong statistical culture when it comes to analysing player performance, few measures have been developed to estimate the potential impacts that recent form may have on a player's underlying ability and expected future performances.
In this talk, a series of statistical models are introduced in an attempt to quantify how player ability and performance might be affected by recent form and other often overlooked variables, including certain match and venue-specific conditions. The validity of the proposed models is evaluated by comparing model predictions of future player performance with forecasts obtained using traditional measures of ability, such as batting and bowling averages. Similarities and differences between the proposed method and other popular player ranking models, such as the International Cricket Council player ratings, are also discussed. A means of combining the newly derived estimates of player ability with a custom-built match simulation engine is then demonstrated, illustrating how the model output can be used to generate probabilistic estimates of possible individual and team outcomes in an upcoming fixture.
Event timings are BST
Pete Philipson for RSS Statistics in Sport Section