Abstract：Galaxy clusters are an excellent cosmology probe because they are in the most massive dark matter halos. The key question with cluster cosmology is how to measure the cluster mass, which could be measured directly by the velocity dispersion of cluster galaxies and gravitational lensing, though both of them are observationally expensive. A more economical way is using the cluster scaling relations. The observational properties of clusters scale with their masses. If the mass is higher, so will be the number (richness) of member galaxies, X-ray luminosity, and the Sunyaev-Zel'dovich (SZ) effect. The scaling relations between X-ray and SZ are consistent with each other, because they are both from the same intracluster medium (ICM). However, a discrepancy emerges when we compare the ICM scaling relation based on SZ data with the galaxy scaling relation based on optical data. To study the discrepancy, we select a complete sample from the two largest bins of maxBCG richness (from SDSS) in the Planck SZ stacking work. In this talk, I will present the X-ray scaling relation of this sample. We confirm the same discrepancy between X- ray and optical scaling relations. I will then discuss that the discrepancy between ICM-galaxy scaling relations may come from mass bias, miscentering, projection, contamination of low mass systems, etc.