Warren Smith, master teacher ;)
===============================

THE GOOD NEWS ON THE FINAL:
---------------------------
According to the computer-grading of the final, only 20%
(that is, 5) of the 25 students in my math55 class who took
the final,  scored below 50th percentile (dept. wide).
(And: the one who scored the lowest basically hadn't been showing
up for class.)

If my students are a random subset of the dept's students,
and I were the same as the other teachers, then the probability
that this few (or fewer) of my students would have scored below
median, would have been
                                             -25           25
( 25C5 + 25C4 + 25C3 + 25C2 + 25C1 + 25C0 ) 2    =  68406/2  = 0.0020.

(Or if you want to use 49% as the chance of being below median, not
50%, then this would have still led to 0.002 to this many decimals 
of accuracy.  And remember, somebody else wrote the final and I never
saw it until the minute you took the test.)

So with 99.8% probability, we're better than they are!
So consequently, our class is going to get better grades than them!
Way to go, team!  Hooray!

THE BAD NEWS:
-------------
You could have done better still!  Quite a few students left early
but made stupid errors they could have fixed if they'd spent that
extra time doing sanity checks and independent recalculations.
Remember, if you have probability P of making an error on a question,
if you do TWO INDEPENDENT calculations of the answer, the probability
you are going to be wrong both times is P*P, which is a LOT smaller.
(If P=0.3 then P*P=0.09.)  And even if you are wrong both times,
the chances you'll get the SAME wrong answer (so you don't notice)
are a LOT smaller STILL.