First, I must thank Shawn for his letter. He has been following TCHCR
since last year (at least) and is probably the most knowledgable head
coach when it comes to the rating. He has even called me to discuss it.
His well-written letter also indicates that he knows what he is talking
about.
Although I have already addressed why the ECAC is rated low in TCHCR, I
haven't dealt with Hockey East in detail. This letter has sort of
inspired me to address the issue, and I am still compiling the
statistics. I hope I can finish it all by Sunday.
(Quotes from Shawn's letter are indented with >.)
>Regarding the computer strength of schedule, there is definitely
>something wrong when the top 18 spots in the computer rating are
>occupied by the 9 CCHA teams and 9 WCHA teams right across the board.
>It is pretty obvious that what has happened is that all of the eastern
>teams have had to schedule their share of independents and therefore,
>their strength of schedule "suffers".
>
>To say that the top eighteen strength of schedules are the western and
>central teams totally skews the college hockey computer rating.....
To get picky, the west-skewed schedules are actually a result of the
west-skewed RATINGS. The average western team is rated much higher than
the average eastern team. Since western teams have almost only western
opponents, their schedules are made up of these high-ranked teams.
Still, there is no denying the western skew. In fact, I have spent many
hours studying it and reporting my findings here to HOCKEY-L. It has to
do with the ECAC's poor performance against the west and the
independent's better performance against the east than the west, among
other things.
>For your interest, and that of your readers, the NCAA does not use the
>college hockey computer rating, but rather a different formula to
>determine "strength of schedule".
The NCAA is using a system, developed for basketball, called the Rating
Percentage Index. The RPIndex is completely different from TCHCR. It
does not use a connected schedule graph but is rather a series of
averages: your winning percentage, the average winning percentage of
your opponents (OWP), and the average OWP of your opponents.
Please note, however, that this western-skew-ness that Shawn mentions
is in the data and not just a product of TCHCR (although TCHCR seems to
magnify it). In the RPIndex, the average schedule rank for the western
(WCHA/CCHA) teams is 12.6; for the eastern teams (HEA/ECAC), 27.3. (As a
comparison, the averages are 9.5 west and 32.5 east in TCHCR.) Using
an even simpler definition for schedule strength, the composite
record of your opponents, the skew continues but is not as wide: 15.6
average schedule rank for the West compared to 24.7 for the East.
>As a coach whose team will be under consideration for an NCAA bid,
>strength of schedule will become important. It is starting to scare me
>that perhaps we might be better off in the future to schedule Canadian
>opponents that have no bearing on our strength of schedule than
>scheduling an independent when maybe they are the only ones a team can
>play. In our league, for example, we only play 21 league games this
>year, so many of us had to fill as many as 13 or 14 non-league games.
>Since the Western and Central leagues had so few openings, we could only
>get so many "attractive opponents", so we just try to fill the rest.
As a teaser for the full report to come, I will slip in some stats
here. Shawn is absolutely right, of course, that HE teams are playing
more independents this year than last year. In the regular season of
90-91, Hockey East teams played 27 of their 96 out-of-conference games
against independent teams (that's 28%). As of 2/9/92, the ratio is
33/76, 43%.
This is not the only change from last year that is hurting Hockey East,
however. If you were following TCHCR last year, you may remember that
Hockey East teams were rated pretty high. The NCAA even sent 4 HEA
teams to the tourney. But times have changed--watch for the full
report.
>Unfortunately, it looks as though that will now hurt us. On the other
>hand, we need to play many of the independents in order to have them
>continue their programs, so we are caught in a catch 22 - it hurts your
>strength of schedule to play a certain independent, but do we want
>college hockey to grow or not?
Shawn raises an intriguing dilemma: what is good for your team may not
be good for college hockey, and what is good for college hockey may not
be good for your team. Which does a head coach serve, his team, or the
sport in general? Shawn could have improved his schedule by not playing
so many independents and simply been idle those weekends, played some
Division III schools or Canadian universities, as he mentions. But
that would not be good for Notre Dame, for example, who got the chance
to play Maine when they were ranked #1 in the polls. I claim that the
interest generated by having UND visit Maine & the experience the Irish
players gained by playing a top-notch squad are just 2 points that make
this type of match-up great for college hockey. I certainly think that,
for the good of college hockey, playing independents and non-conference
games should be ENCOURAGED and not discouraged.
But what's done is done for 1991-92 and we must live with it. The NCAA
cutbacks from 38 to 34 games (with 6 months to implement the change!)
really screwed up the scheduling. Let's look to the future, however,
and implement some changes so Shawn is not forced to schedule Canada
Tech.
I have been playing around with an idea for many months now, an idea I
call "cooperative scheduling". It involves having the different LEAGUES
work together to schedule games. Right now, the leagues only take care
of scheduling the league games. Each individual school is responsible
for doing its own non-conference games. This can lead to some pretty
lopsided schedules: this year only 1 game between the WCHA and the ECAC
will be played and the eastern teams were forced into playing more
independents.
Cooperative scheduling means the leagues need to set up agreements with
each other to schedule games. For example, the CCHA and HEA may say,
"Every year, we will play at least 8 games against each other". The
leagues would help organize all of this, with perhaps the CCHA office
calling Miami and Ohio State to see if they would want to host BC and
BU. And as a reciprocal, maybe BG and Western Michigan could travel to
Merrimack and Lowell.
Included in all this, of course, would be the independents. If each
league would agree to play 10 games versus the independents, then the
problem that Shawn has identified would not be so severe.
And each team would still have complete control over some of its
non-conference games, maybe 2 of them. Harvard would undoubtedly use
its 2 for the Beanpot, but Vermont could choose to re-new its rivalry
with New Hampshire, or whatever.
With 3 independents joining the CCHA, and Anchorage possibly getting in
the WCHA, this may indeed be a great time for the leagues to get their
acts together and start cooperating with one another.
Just weeks ago, I heard on HOCKEY-L about an agreement between
Clarkson, St. Lawrence, Maine and Providence. This is wonderful, and
we need more of it. I think that by getting the leagues involved, it
will be easier.
Also, I have been helping the CCHA with their league scheduling for the
next 2 years. While we were working at it, we tried to look at the "big
picture", too. For example, we stayed away from placing league games
over the Thanksgiving break. If all of the leagues agree to not play in
their league over that period, then all sorts of combinations of
non-conference play can happen then. Also, when 2 CCHA teams were going
to be idle from league play over the same weekend, we tried to pick 2
teams that were geographically close to each other, like Michigan and
Michigan State. Then, it would be possible for two teams, maybe North
Dakota and RPI, to visit, each playing Michigan one night and MSU the
other. This type of scheduling adds "connections" to the schedule graph
and makes for better comparisons. It is good for college hockey!
I see some problems right off the bat. The WCHA must change from its
32-game league schedule or else they will not be able to participate in
this cooperation. The CCHA teams will have 4 NLGs (unless Ferris drops
out), which may not be enough. The Ivies might have to agree to play a
few more games. HEA is already set with 10 NC games, by the way.
This is the first I have told a large number of people about my idea.
Let me know what you think.
Keith
|
