Me too! But I think it comes down to the fact that there's far more involved here than just numerical specifications. Roger Clark's evaluations are based on the specs - with pixel sizes, read noise and well depth in terms of number of electrons, etc., used to drive mathematical models of sensor performance. The DxOMark score is a composite of ISO sensitivity, bit depth, and dynamic range. If you want to isolate ISO sensitivity, click the Sports tab. In that category, Nikon still comes out on top, but all of the top 10 for that list are FF dSLRs, and the Phase One MF cameras don't show up until #11. Even so, if you look at the top two sensors, there's a substantial difference (ISO 3250 vs. 2300, so ~1/2 stop) between the D3s and the D3 - and both have the same size sensor and the same size pixels! Clearly, there's more to the story...