I read the article you linked to but didn't find it convincing. It seems to be saying, "I was a teacher who failed to have much effect on poor students. There

I read the article you linked to but didn't find it convincing. It seems to be saying, "I was a teacher who failed to have much effect on poor students. Therefore, teachers cannot have much effect on poor students". We seem to disagree about how much impact a teacher can have on poor students, but I wonder if there is an experiment we can agree on which will settle the matter. For example, imagine there are 400 schools which have two math teachers, and that the students are assigned to one of the two teachers at random. Now imagine that each year we give an award to one (out of the two) teachers whose students performed best. If teachers have no impact on student outcomes, we would expect that random chance means after 3 years about 50 teachers have won the award three times. But if we find that 125 teachers have won the award three times wouldn't that suggest that teachers have significant impact on student outcomes?