I develop physical models and statistical methods which allow the data from large scale galaxy and cluster surveys to constrain models of galaxy formation and cosmology.

I have played a leading role in the development of what is now the standard model of nonlinear clustering and biasing: the Halo Model. It is the currently the best language for interpreting measurements of weak lensing, the thermal and kinematic Sunyaev-Zeldovich effects, and how galaxy clustering depends on galaxy type, both in real and in redshift space. In 2004, I showed that dark matter halo formation is correlated with environment; I also discussed why, and pointed out that understanding this correlation is necessary if the Halo Model is to be used as a precision tool for cosmology. This correlation was sufficiently unexpected and the consequences sufficiently important that a number of groups have since confirmed that the effect does indeed exist: the effect is now called Assembly Bias. I have also used physically motivated models to illustrate the ubiquity of what are now called scale-dependent bias, tidal bias and velocity bias.