The thing with 500 MB Heights, is that small differences in scores actually could represent rather important differences with spatial outcomes. It may look statistically insignificant, but those differences could represent small differences in placements of ridges and troughs at Day “X” but the sensible wx outcome for a specific point could vary quite a bit based on these differences. In general at day 3 we expect all models in general to show a trough or ridge in the same general place. However a small change in 50 miles of the 500mb 560 height contour, may matter significantly with downstream sensible wx outcome.