The point of the question is that the conclusion doesn’t necessarily follow logically from the premises — we are trying to find a principle that is used implicitly in the argument. And the argument isn’t talking about how probabilities transfer from premises to the conclusion.

(B) means that a conclusion is undermined when what we would expect to happen if the conclusion was true doesn’t actually happen. (B) says that if we expect that a reshuffle would be preceded by a meeting, and a meeting doesn’t happen, a reshuffle is less likely.

Here, the hypothesis that the premier will reshuffle the cabinet isn’t supported by all the available data. In fact, the author gives reasons that the hypothesis is unlikely to be true. So (C) isn’t relevant.

The argument is trying to conclude that it is proper, not erroneous, to assume a phenomenon will recur only under the circumstances in which it previously occurred. Additionally, the phenomenon (reshuffle) wasn’t necessarily caused by the circumstances (meeting) in the past.

This is true, but it’s not used in the argument. We’re not given 2 statements that are inconsistent with each other.