For hypotheses about “some completely novel thing,” it’s usually the case that there’s an existing narrative about what that thing would be like were it to exist. 

Observations are then cherry-picked that seem to support the existence of that expected thing. 

But because the thing doesn’t actually exist, the data are at best highly ambiguous, unclear, ambiguous. 

For example, the thing might be
- Big Foot
- UFOs
- Covid
- Climate change 

All have the hallmarks of this. We find the “data” backing exactly what we expect such a non-existent thing to be like, and the data are inherently poor. 

Cultures seem to select for the constant acquisition of “evidence” for such chimeras.