Discriminant Analysis Issue

I'm trying to make sense out of a recent discriminate analysis I ran which confused me. I got really a really low eigenvalue which translated into a very low canonical correlation score. The Wilks' Lambda turned out to be significant however. So basically, I got a weak model but it is significant. That confuses me a bit, any help on this would be appreciated. But the main issue I have is that if the model is weak how come that isn't translating into the classification results being weak as well?

Eigenvalue: .143
Canonical Correlation .353
71% cross validated grouped classified correctly

This were 13 independent variables and the dependent variable had 2 options which results in 1 function being produced.