Wikipedia says it nicely:
> the "informational value" of a communicated message depends on the degree to which the content of the message is surprising [0]
Of course, this assumes the message is factual. If it's not factual then the message carries no information (aside from the information it implies about the sender). But, that assumes the receiver is able to discern whether or not the message is factual. If the receiver can't discern that, then it simply appears to be a highly informative message.
False outrageous claims have the further advantage that they are unconstrained by reality. True surprising claims are by nature less likely to occur. That's why they're surprising. But there is no probabilistic throttle on false outrageous claims - one can generate them at whatever frequency they desire.
Thus, false outrageous claims have an information theoretic advantage over true claims because they outnumber true surprising claims, and they appear to be more informative than true unsurprising claims. My argument isn't that the more outrageous a claim is, the more likely people are to believe it, per se. My argument is that outrageous claims get more attention from people because they are less likely, and those who are unable to discern that they are false will simply see them as highly informative.
You could probably argue further that the more frequently an outrageous claim is repeated, the less improbable it will sound. In other words, start with an outrageous claim to get peoples attention, and then repeat it a lot to normalize it and reduce doubt about it's validity.
0. https://en.wikipedia.org/wiki/Entropy_(information_theory)