Comment by joe_the_user

I've heard Naives Bayes described as "nearly good enough" for many uses. Can anyone quantify how much worse it is for various applications?