I look at the accuracy number and if it is high I think we are fine. But there are all these other metrics I do not really understand, and I could not tell you when my model is actually failing. I nod along in the meeting and quietly hope nobody asks me to explain.
More people experience this than they realize.
You only understand accuracy and cannot tell when your model is truly failing.
If this sounds familiar, the Library can help you find the bigger picture.