It would be interesting to look at distribution of death/injury by days post vaccination, for both high and low-injury lots. I wonder if the distribution of AEs for low-injury lots is random (implying the vaccine was not causal), while the high-injury lots show high rates of AEs shortly after vaccination (implying causality).

Rose did an analysis of death/injury by days after vaccination, implying causality: https://cf5e727d-d02d-4d71-89ff-9fe2d3ad....