Halos have an edge, and the measured edge is in the wrong place. The splashback radius, where accreted matter reaches the apex of its first orbit and the density profile steepens sharply, is one of ΛCDM's cleanest structural predictions: its location, near 1 to 1.5 times R200m with a known dependence on accretion rate, follows from collisionless dynamics with essentially no baryonic ambiguity. Measurements found the feature, a genuine success, but displaced: cluster-galaxy cross-correlations around SDSS redMaPPer clusters located the steepening at radii about 20 percent smaller than simulations predict for halos of the same mass, a result echoed in DES analyses and visible in weak-lensing profiles, with the discrepancy's significance debated ever since.
The defense has focused on the cluster-finder: optical selection with a fixed aperture preferentially picks concentrated systems and imprints aperture-scale artifacts onto the inferred profiles, and SZ-selected samples (ACT, SPT) yield splashback radii closer to expectation, though with larger errors and their own selection geographies. Yet the optical-selection account must explain a 20 percent dynamical displacement through percent-level selection weights, mass-mixing corrections leave residuals in several reanalyses, and the galaxy-population dependence of the feature (redder galaxies showing different splashback locations than blue) adds dynamical structure the artifact reading does not predict. The feature is real, measurable, and persistently smaller in the best-measured samples.
The standing is an unresolved precision discrepancy in the one halo observable that was supposed to be theoretically pristine: Euclid and Rubin will measure splashback in lensing alone, selection-independent, and settle whether the edge sits where collisionless dark matter puts it.