Erik writes:
> mike writes:
> >              Without the numbers here, I would estimate that a
> > pretty high percentage of WCHA/CCHA teams who won the first game did not
> > win the 2nd.  Personal experience of seeing back-to-back games between
> > the same two teams also makes me believe the Game 1 loser fares better
> > the second night.
>
>  Hmmm, I guess I'd like to see some numbers.  After watching 11 seasons of
>  2-game series in the WCHA, my impression is that if the HOME team loses, it
>  will very likely come back and win the second game.  If the visiting teams
>  loses the first game, it seems quite likely to lose the second as well.  It
>  is very difficult to sweep a 2-game series on the road in the WCHA, but not
>  so unusual to do so at home.
 
I agree that it is always difficult to sweep on the road, outside the WCHA
too, so I guess the question is what happens when the visitors lose the
first game - what percentage lose the 2nd as well?  I'll try to dig up
some numbers at home, please give me some time.  Anyone else who wants to
do so, feel welcome.
 
I also want to add to what I said originally.  I believe that even teams
that are swept, are likely to lose the 2nd game by a smaller margin than
the first (and so they usually play "better").  Of course I have no
statistical basis for this at the moment, it's just an impression I have
and have talked about with other people.  This would show up as being swept,
but it could be considered in favor of the losing team if they play a
better game the second night.
 
This brings up an interesting sidepoint that indirectly deals with the
new NCAA format.  Does a single game tend to be won by the "better team"?
And if they do win, are they likely to win or lose the 2nd of a series?
I believe that a single game affords the underdog a better chance to win
than a three-game set.  But what about two games?  Is an underdog likely
to split 2 games?  This is difficult to determine in regular season records
because it involves going back and setting a "favorite" for each series,
which may or may not be a home team.  Suffice it to say that most visitors,
even if they are a better team than the home team, are satisfied to escape
with a split.
 
So, when I do my survey of past two-game sets, I'll be ignoring any
idea of a "favorite" and just concentrate on home/visitor and also margin
of victory.  I've got DivI scores going back 5 years so this should be
a pretty good sample.
 
 
- mike (taking on more work for the good of the list...)