That "common knowledge" is wrong. Myers briggs is not great for scientific studies since it splits people into 16 buckets instead of sliding scales (like the Big 5, which is used in research). But MB is an incredible mental toolkit to have.

Why would something with no evidence behind it be an incredible mental toolkit?

See the whole criticism section.

https://en.wikipedia.org/wiki/Myers%E2%80%93Briggs_Type_Indi...

MB is astrology for nerds.

Astrology has a one in twelve chance of accuracy since the only variable is birthday.

With the four discrete variables in MBTI, I can type people with 90% accuracy after a conversation.

The buckets themselves are also false, so no.

Isn’t this often just the Barnum effect, especially when ground truth is established by asking people if you are right? https://en.m.wikipedia.org/wiki/Barnum_effect

Not quite zero substance, the four dimensions are reasonable close to some of the five leading components of personality.

The problem is that each letter comes from splitting a normal distribution in half right at the middle where maximum likelihood is. So most people won't be very strongly attached to most of their letters.