I find that weird outliers are usually evidence that the underlying model is flawed in some fundamental way, but sometimes I think people believe the exceptions

I find that weird outliers are usually evidence that the underlying model is flawed in some fundamental way, but sometimes I think people believe the exceptions are fundamental, and a lot of effort goes to the wrong place.