If you take reactions out of their given context, then this holds: without knowledge of the situation, all the given parties are indeed equally likely to be ove

If you take reactions out of their given context, then this holds: without knowledge of the situation, all the given parties are indeed equally likely to be overreacting. But real-world situations don't work that way. They bring their own context, and in doing, so they break the symmetry: unless all parties are reacting in the same way, it becomes possible to determine which party is more likely to be overreacting to that situation, often with shocking ease. From there, of course, one can and should dig deeper. It's in the nature of probability that sometimes, reality will go against the likely outcome. But it's also in the nature of probability to hold often enough to be useful.