I am so entirely amused that (a) you actually tabulated the data (because of course you did), and (b) there does seem to be a chance there's a non-negligible signal there...