Recent studies of galaxy clusters has provided loose evidence for the existence of self interacting dark matter. These measurements on single cluster environments and major mergers have suffered from assumptions on the time of in-fall, orientation of impact and line of sight projections. In this talk, I will present a unique method to measure the interaction cross-section of dark matter using an ensemble of galaxy clusters, which is independent of all these factors. Using hydrodynamical simulations, I will show that one should be able to constrain the cross-section to extremely high accuracy using only 40 clusters, a typical sample size one would expect from the Hubble archive. I will finish by commenting on how one would be able to extend this theory to missions such as Euclid and LSST.