I'm not sure how to describe this in rigorous Bayesian terms, but things usually happen in patterns. When we make inferences, we cannot ignore the context.
Based on the recent history of violence in France, is it more likely that a mass killing committed by a Franco-Tunisian (who are mostly Muslims) on a clearly suicide mission is related or is not related to islamic terrorism? Better yet (since money does have a tendency to clear the mind), what would you bet $100 on?
I'm not saying that we don't need facts. We still have to confirm everything, and strange unlikely things do happen. But at the same time, we cannot act as if we are completely ignorant until we get substantial evidence.
There's not much evidence yet, but also not much doubt.