Clusters of galaxies are great laboratories for many astrophysical processes on galaxy scale and have become one of the key probes in today's cosmology. Especially their mass function redshift distribution helps unveiling the effect of the mysterious driving force on the expansion of the universe - dark energy. In order to step forward to even higher precision cosmology using clusters, one not only needs to build good analysis tools, such as cluster finders, but also understands their systematics.

In this talk, I will map out a round trip from understanding systemactics in large cluster samples to improving those tools for better cosmology. I will present a way to improve the understanding of the Universe as an interlocking element between real data and simulations. This will include, 1) building an empirically-motivated mock catalog using high-resolution N-body simulations to test various analysis tools, 2) understanding contamination and incompleteness using cross match of multi-wavelength cluster samples from large sky survey data. These projects are crucial in interpreting data from cluster surveys, such as South Pole Telescope (SPT), Dark Energy Survey (DES), as well as an upcoming survey from the Giant Magellan Telescope (GMT).