Because it's very hard to rationally compare multimodal batches without single test statistics. And they present five summary figures for each batch, each of wh

Because it's very hard to rationally compare multimodal batches without single test statistics. And they present five summary figures for each batch, each of which are reasonable metrics to compare batches with.