One of the upshots of a number of my posts on the limitations of probability theory is that there are events that are probabilistically incomparable—neither can be said to be more likely than the other. (For instance, this post.) But an objective chance at a good is good, and better the greater the chance and worse the lower the chance. Chances p and q at the same good G will, then, be incommensurably good when the chances p and q are incomparable. Hence, if there are incomparable objective chances, there can be incommensurable goods. But it's plausible that there can be incomparable chances (see the post I linked to above, for instance). So there can be incommensurable goods.