# no graphics; Who Did What consisted just of the word "stuff"; blythe use of testing; inadequate explanation of external methods--chi-squared, naive Bayes, SVM (especially SVM whose description was not only skimpy but misleading)); did notice that using NO predictors in bank acct problem is just as good; explanation of CIs in Problem C flagrantly wrong

pts to make in blog:

1.  most groups blythely used sig tests!

2.  excellent writing, graphics etc.

3.  it was common for people to bring in external methods but not
explain them, as required

4.  Problem C:  article's example of power doesn't make sense; the power
stuff sounds quasi-Bayes [P(H0) = 0.1), but since Bayes Rule is not
sued, no harm no foul; Higgs--tails inaccurate, as I hinted in 3-4
lectures; article misses the fundamental problems with testing

5.  using "I" means the group didn't review it; not professional