Parameter biases of the data vectors with the original and ITEM priors. The red and blue ellipses show wide contours (same from Fig. 6), while the yellow and green ones were obtained after enforcing the 0.4σ_2D bias threshold and 1D coverage threshold for the 2D marginalized constraints. Both are centered in their respective 2D marginalized MAP. Due to parameter volume effects, the marginalized constraints from the baseline prior analysis are not centered on the input cosmology. Top: simulated likelihood analysis for the data vector without shot-noise (α₀ = 1.0, α₁ = 0.0). Bottom: simulated likelihood analysis for the data vector with Buzzard stochasticity (α₀ = 1.26, α₁ = 0.29). The dashed horizontal and vertical lines indicate the fiducial parameter values.