Points per tossup has some potential variance due
to differing moderator reading speed. A faster
moderator gets to a power point faster, so a player must be
faster on the buzzer to get the power, compared to being
read to by a slower moderator. In a tournament set
with many powerable tossups, or in one with a great
diversity in moderator speed, this has a greater potential
to be a problem.
Of course the faster
moderator tends to be considered the better moderator. Some
TD's I have talked to about tournament schedule design
will manipulate things so that important games between
top teams will be played in rooms with the
better/faster moderators. A simple powers/tossups read stat for
a moderator would be a bad test to see if this is a
significant problem.
Anthony, whose training in social
science research methods suggest a more complicated
method probably worth too much effort to see what the
actual effect, if any, is.