Overtime:  10 minutes vs. 5 minutes
 
Patrick Blake's post, about the time goals are scored in overtime,
prompts this post.  For the purists out there (who count me as their
number), statistical evidence seems to suggest extending overtime to
10 minutes if you want a decision in a game without using a shoot-out.
This will not, however, avoid ties altogether.
 
Theory:
 
A 10 minute overtime will yield more won/lost decisions, since there is
more time for teams to score.
 
Data Background:
 
RPI overtime statistics since 1949 (when the program was re-started).
Yes, there was overtime prior to the war, but there were no standard
rules, at least in RPI games.
 
This yields 122 overtime games played by RPI.  Of this, 9 must be
eliminated, as they occurred during tournament competition, therefore
there was no time limit imposed on the overtime.  It was simply "next
goal wins."  From a tactical standpoint, this will make a difference in
how teams play.
 
Consequently, we are examining 113 games.  Since RPI is in the
ECAC, 10 minute overtime periods were played through the 1988-89
season.  This results in 88 games played with 10 minute overtimes and
25 with 5 minute overtimes.
 
Results:
 
From 1949 to 1989, 69 of the 88 (78.4%) overtime games resulted in a
decision. 19 (21.6%) ended in a tie.
 
From 1989 to the present, 13 of the 25 (52.0%) overtime games
resulted in a decision. 12 (48%) ended in a tie.
 
On the surface, doubling the overtime period cuts in half the number of
ties.
 
To delve further, I checked how many of those 69 games resulting in a
decision would be ties after 5 minutes.  The total was 31 (44.9%) of
the 69.  To put it another way, retroactively applying a 5 minute
overtime to those games, 38 (43.2%) would have resulted in a
decision, 50 (56.8%) would have ended in a tie.  These revised
numbers are closer to the current 5 minute overtime than the current
overtime is to a 10 minute overtime.
 
Conclusion:
 
A 10 minute overtime would yield fewer ties.
 
Points to consider:
 
1)  Retroactively applying the 5 minute overtime to games is a
somewhat shaky maneuver.  How many times did a team pass up a
marginal scoring opportunity in the first few minutes, knowing there
was plenty of time?  With a 5 minute rule, the team might well go for
the scoring opportunity.  These situations cannot be considered
adequately.
 
2)  This uses only RPI games.  I have no other statistics to use, though
I suspect the results are not much different for college hockey as a
whole.
 
3)  I did not run a statistical analysis of the data to verify the
hypothesis, for five reasons.  First, the resulting numbers satisfied my
curiosity.  Second, I did not want to bore the non-mathematically
minded people on the net with the details of the analysis.  Third, I did
not want to bore the mathematically minded people on the net with the
details of the analysis.  Fourth, I did not want to increase the length of
this post.  Fifth, and most importantly, I am lazy.
 
Finally:
 
I have had overtime data for a couple of years and knew the reduction
in the length of the overtime period did increase the number of ties.
Without resurfacing the ice, an extra five minutes would only add 10-
12 minutes to the length of a game.  That's a small price to pay for
cutting the number of ties in half and teams "kissing their sister" less
often.  Poor ice would be a disadvantage to both teams.  In my
opinion, it is a more attractive option than the shoot-out, which adds a
similar amount of time to the game, and does not match in excitement
to sudden-death overtime.
 
Kurt Stutt
[log in to unmask]