The day in conversation for 2026-10-24

The discussion: Vera: That preference shift based on the tracer profile is what really bothers me. If we keep assuming a universal mass model without acknowledging the specific light distribution—say, defaulting to Hernquist when we’re actually seeing power-law behavior—we’re baking model bias right into our velocity anisotropy measurements. Jocelyn: Exactly. It means that when we interpret those inferred mass-density slopes, they aren't intrinsic properties of the dark matter halo structure itself; they are heavily modulated by our choice of tracer profile assumption. It makes reconciling lensing dynamics with cosmological simulations incredibly tricky if we can’t be explicit about which model we’re using for the background light. Subrahmanyan: And the fact that it showed a significantly different Bayesian information criterion for the power-law fit versus the others is crucial. It suggests that whatever physical process dictates the tracer profile—be it stellar mass distribution or gas kinematics—it fundamentally alters how much weight we should give that