[richrichie]

Indeed. The fundamental insight behind CLT i.e. sample average is normally distributed even when population distribution is not normal is intuitive, yet the the theorem is magical.

[hinkley]

When you have more than one variable, the outcomes get clumpy.