I outline four competing probabilistic accounts of contrastive evidential support and consider various considerations that might help arbitrate between these. The upshot of the discussion is that the so-called 'Law of Likelihood' is to be preferred to any of the alternatives considered.