Abstract: The study of cosmology is entering an era of higher quality and higher quantity era that is moving us from being limited by the raw statistics of our data to instead be limited by the accuracy of our underlying physical models. A specic set of models where attention of this problem has been growing is the use of halo models to connect the invisible world of dark matter to the visible world of stars and galaxies. Most halo models make their predictions by taking the mass of dark matter halos as the sole parameter. The literature has demonstrated that secondary halo properties can have enhanced clustering compared to the general population of dark matter halos, in an effect that is referred to as \halo assembly bias." Neglecting halo assembly bias from our models can result in severe biases. I add to the literature by making the rst detailed study on the choice of halo denition on common measures of halo assembly bias. I utilize non-traditional halo denitions seeking a choice that minimizes the impact of environmental effects which may drive halo assembly bias. I nd that halo assembly bias is a strong function of halo denition for the properties of halo concentration, halo shape, and halo spin. I demonstrate that the impact of halo redenition is primarily caused by the changing host halo populations, as neighboring halos are demoted to substructure. I further show that these results are consistent with those of the \halo splashback radius"; however, halo splashback radius does not increase halo sizes suffciently to remove halo assembly bias for most scales or masses. I discuss how these results give us insight to the relevant scales of what might be driving these relations and how they give a better understanding of galaxy formation and galaxy evolution. I conclude by laying out a course for the future with multiple paths to better understanding halo assembly bias and constraining how it impacts the models as well as utilizing it as a probe of galaxy formation.