A dot plot that shows samples of the same type generally had similar levels of error on individual benchmarks.