I reviewed the data compiled by the Broad Foundation and suspect that CMS won based on two metrics. The first was a residuals analysis, a statistical measure that allows one to compare predicted and actual performance. The residuals analysis determined that CMS students performed better than expected on state tests, i.e., positive residuals, relative to other districts in the state.
The second was the percentage of achievement gaps (as measured by state test results) closed. The math and reading proficiency gap between white students and their black and Hispanic classmates closed by one or two percentage points, but the gaps themselves remain pretty large. For the Broad analysis, the size of the gaps appeared to be less important than the number/percentage of gaps closed over the last three years.
Of course, these measures require one to have a great deal of faith in the state testing program.