If you want to understand this issue please take the time to look at the actual MET project report as well as other well done research. Gary Rubinstein's analysis seems very misleading (see comments to prior HN story). Apparently is is very controversial to even try to measure teacher effectiveness.