how to interpret missing probability in goodness of fit test when running a MNL model?

[Tom Thumb]

I just ran an MNL model regression in STATA 18. output is good.

I am trying to run GOF using

mlogitgof, table

command in STATA 18.

I am getting the following result. How do I interpret this? It looks like something is wrong.

Please help. Thanks so much.