The NCAA reduced the length of overtime in regular season
games from ten minutes to five minutes for 1989-90, and
since I have all results for 1988-89 and 1989-90, I thought
I'd do a quick study and see if the shorter overtime had
any discernable effect on the number of games that ended
up as ties.
Of course, including several more years prior to 1988-89
might give a different result, as would including more
years with the 5-minute OT, but I don't have easy access to
either of those sets of data. I also do not have the times
that the game-winning goals were scored in most of these
games, otherwise I'd like to see just how far into OT a
game-winner tends to come.
The games under consideration were all regular season games
played between two Division I teams in each of the two years.
Division I teams were considered to be teams in the four
conferences, plus the following independents: Air Force,
Alabama-Huntsville, Alaska-Anchorage, Alaska-Fairbanks, Kent
State, Notre Dame, and St Cloud. Since Kent did not play in
1988-89, there were a total of 43 teams considered to be
Division I in 1988-89 and 44 in 1989-90. Merrimack was listed
with independents in 1988-89 and with Hockey East in 1989-90.
Since the objective was to determine the overall effect of
shortening overtime, it did not really matter what group a
team was placed with except when looking at the league numbers.
There were a few games over the two years that were actually
decided in double or triple overtime in the regular season
(tournament games). I considered a game a tie if it was not
decided until after the first five or ten minutes of overtime
(depending on which year). Also, a game in the 1988 Nissan/Jeep
tourney in Alaska between Michigan Tech and Minnesota was tied
at two after regulation, but since it was the consolation game,
no overtime was played (this does not happen in all tourneys).
So, I threw that game out.
The following charts indicate 1) what percentage of games went
into overtime by league each year, and 2) what percentage of
overtime games were decided, i.e. there was a winner. The
abbreviations are: GP - number of games played (in a league,
for example), #OT - number of those games that went into OT,
%OT - percentage of those games that went into OT, #dec. -
number of overtime games that were decided, %dec. - percentage of
overtime games that were decided, #T - number of OT games that
ended up tied, and %T - percentage of OT games that ended up tied.
Leagues are: CCHA - Central Collegiate Hockey Association, WCHA -
Western Collegiate Hockey Association, ECAC - Eastern Colleges
Athletic Conference, HE - Hockey East, Indep/NL - games played
between independents or nonleague games played by league members,
and HE-WCHA in 1988-89 refers to the 56 (8x7) games played between
members of those leagues as part of the interlocking scheduling
agreement that terminated after that year. Those games counted
in the standings of both HE and the WCHA.
1988-89
-------
League GP #OT %OT #dec. %dec. #T %T
===========================================
CCHA 144 30 20.8 13 43.3 17 56.7
WCHA 112 25 22.3 18 72.0 7 28.0
ECAC 132 15 11.4 11 73.3 4 26.7
HE 63 5 8.9 2 40.0 3 60.0
HE-WCHA 56 9 16.1 5 55.6 4 44.4
Indep/NL 161 18 11.1 9 50.0 9 50.0
=================================
Total 668 102 15.3 58 56.9 44 43.1
1989-90
-------
League GP #OT %OT #dec. %dec. #T %T
===========================================
CCHA 144 24 16.7 10 41.7 14 58.3
WCHA 112 14 12.5 9 64.3 5 35.7
ECAC 132 22 16.7 10 45.5 12 54.5
HE 84 13 15.5 7 53.8 6 46.2
HE-WCHA -- -- ---- -- ---- -- ----
Indep/NL 216 24 11.1 13 54.2 11 45.8
=================================
Total 688 97 14.1 49 50.5 47 49.5
Here are the totals again so we can compare them:
GP #OT %OT #dec. %dec. #T %T
1988-89 668 102 15.3 58 56.9 44 43.1 10-min OT
1989-90 688 97 14.1 49 50.5 47 49.5 5-min OT
Just by looking at these two years, it looks like
the shorter OT has resulted in more ties, but not
too many more. In fact, the overall rate of ties
divided by games played only increased from 6.587%
to 6.831%.
Leagues also seemed to pretty much maintain their rates
of ties and OT games. The ECAC did have a larger rate
of ties this year, but that was really an exception
and there's not enough information to determine if that was
a result of the shorter OT.
So, even though I preferred the longer overtime (hey, I
wish every game was sudden-death!), I have to conclude that
there's no real evidence that a shorter overtime results
in more ties. At least, not significantly more.
There you have it. Enjoy it, show it to your friends,
whatever. I'll leave you with this thought:
Ohio State has not lost any of their 15 regular season overtime
games (5-0-10) in the last two seasons. The Buckeyes are
20-46-10 in that time.
- mike
p.s. there's no truth to the rumor that I started this study this morning
at 2 am after watching Edmonton beat Boston in triple overtime...
|