On Feb 3, 10:51, "Greg R. Berge" wrote (about them ECAC tie-breakers):
>3.  Once upon a time, # of wins was one of the criteria.  For example, a
>13-9 record would beat out a 12-8-2 record.  I didn't know this had dropped
>out of consideration (though I'm pleased).
 
The odd thing about this is that, while it hasn't been a criterion in ECAC Div.
I as far as I know for at least the last eight years, # of league wins has been
and continues to be used as a tie-breaker for the ECAC's Div. III leagues.  It's
used either just before or just after head-to-head record.  Other than that, the
tie-breakers at the ECAC's Div. I and Div. III levels are identical.  Odd that
the leagues choose to differ in this way.
 
>4.  Does it seem strange to anyone that while goal differential is
>enshrined in this system as so important, goal differential in ALL RS games
>is never considered?
 
The whole Top 4/Top 8 distinction seems strange to me.  Top 8 especially, since
there are now 10 playoff teams.  But yeah, if you're going to insist on using
goal differential as a tiebreaker (and I wish they wouldn't 'cause I hate it,
but that's a whole nother tangent), it makes sense to consider all games, since
all games are considered when you look at the teams' league record... which
presumably is leading to the tie in the standings that's creating the mess in
the first place.
 
I don't believe the ECAC has ever had to go past the third tiebreaker, so goal
differential (however they want to use it) appears never to have been a factor.
 
>5.  It seems to me that these rules can produce an infinite loop with no
>resolution when there are ties at the 4/5 and 8/9 splits.
 
Yup.  Should this happen, both the 8/9 preliminary-round game and the 4/5
quarterfinal series would be held at neutral sites.  :-) :-)
 
(The situation should be statistically rare, but not impossible.  I suspect that
if the ECAC does encounter something like this, they'll make up another rule on
the fly, possibly involving goal differential in all RS games :-)
 
>I think this problem goes away if you remove any reference to
>seeding within the seeding rules (since this is the source of the circular
>reference).  You can do this by saying, for instance, record vs. teams >
>.500.
 
Sounds good.  Please write Joe Bertagna at...
 
--
Disclaimer -- Unless otherwise noted, all opinions expressed above are
              strictly those of:
 
Bill Fenwick
Cornell '86 and '95
LET'S GO RED!!                                                  DJF  5/27/94
"So what's Cornell's strategy now -- go to the net?"
-- Grady Whittenberg, Cornell (WQNY) play-by-play announcer, following the
   clean-up efforts after Providence goalie Dan Dennis got sick on the ice
   during the first round of the Syracuse Invitational
 
HOCKEY-L is for discussion of college ice hockey;  send information to
[log in to unmask], The College Hockey Information List.