Interest in using redshift space distortions (RSD) as a precise test of General Relativity on large scales has exploded in the last few years, but our theoretical models must be improved to match the statistical precision of current data. In this talk I will review of physics of redshift space distortions in configuration space and present projections for RSD constraints from the BOSS survey. While many recent studies focus on understanding dark matter clustering in redshift space, galaxies occupy special places in the universe: dark matter halos. Our model reproduces the complex dependence of redshift space clustering on halo bias seen in N-body simulations and is sufficiently precise to model the final BOSS data set on scales of r >~ 25 Mpc/h. A better treatment of 'fingers-of-god' from satellite galaxies is still needed to maximize the cosmological information available from RSD on these large scales.