Those rankings are idiotic. Except for the complete cluster17 states, mostly they just give you a rough order of the demographics of each state. The rankings look a lot different if you compare students of comparable backgrounds from each state.
And I'm not sure what methodology they're actually using, but the statistics they list are pretty much all indirect measures at best of quality. Spending per pupil is not very strongly correlated with performance and graduation rate, while helpful, doesn't mean much if there is no standard assessment test enusring they graduated because they obtained the necessary skills and not because the schools just move people up and out.