A Simple Upper Bound on Overtime Scoring

September 17, 2015


The number of NHL games requiring an overtime has been fairly steady since the 2004 lockout. On average, 76% of NHL games are decided in regulation, meaning about 20 games of a team's 82 game schedule will need extra time.

Under the current NHL overtime format of five minutes of 4-on-4 hockey, 44% of overtime games are decided by a goal scored in overtime. Given that the average NHL team plays in 20 overtime games, nine of these games will see a goal scored during the overtime period. A typical NHL team can reasonably expect to score about 4-5 overtime goals per season.

The new format for an NHL overtime will see teams play 3-on-3 hockey. It is anticipated that more of these overtime games will be decided by an overtime goal (instead of the shootout). The amount of overtime games ending with an overtime goal is the cause for much debate, the source for which is the uncertainty of goal scoring rates for 3-on-3 hockey. Because there is very little data for 3-on-3 goal scoring rates in the NHL (approximately 90 minutes worth over the past decade), estimates vary dramatically depending on the sample size used in the study.

Determining the Upper Bound

Despite this uncertainty in 3-on-3 goal scoring rates, it is easy to compute an upper bound on overtime scoring under the new overtime format. Imagine that every game requiring overtime is decided by an overtime goal; that is, there are no shootouts in the NHL in 2015-2016. It is a decidedly wrong assumption, but it allows us to determine the maximum increase in goal scoring due to the new overtime format. Instead of nine of the 20 overtime games ending with an overtime goal, all 20 will be decided before the shootout. And so, instead of nine goals being split between two teams, there will now be 20 goals.

The average number of overtime goals will increase from 4-5 goals per team (per season) to 10 goals. With an increase of about five goals per team comes an increase of about eight assists (on average, 1.7 assists are awarded per goal in the NHL). Thus, in the most extreme scenario of all NHL overtime games decided before the shootout, the average NHL team can expect an increase of approximately 13 total points (goals plus assists). With a typical NHL team scoring 591 points per season, this extreme example projects a 2% increase in the overall point totals for NHL teams.

A More Realistic Approach

A more techincal calculation of the increase in scoring due to the new overtime format was performed in our fantasy hockey draft kit (we use a range of 3-on-3 goal scoring rates to determine what fraction of NHL games will end in overtime under the new format). The result of this calculation pegs the (per-team) increase in scoring closer to 1% (about 7-8 extra points per team). The important part of that sentence is the per-team phrase. These extra points still have to be divided up among the team's three player overtime units.

If you're interested in seeing how the new overtime format and the new face-off rule will impact your fantasy hockey team, be sure to download a copy of our fantasy hockey draft kit. These rule changes are also the topic of our most recent fantasy hockey podcast.


will be interesting to see how teams line up for OT - go for broke w/ defensive minded C and Ws or put an offensive minded D out there with 2 forwards, or even go for the tie and play 2 Ds ?

@george8kl Yeah, most of the fun will be in first quarter of the season as teams try to figure out what works best.

Incidentally, the composition of the 3-on-3 lineups is one of the reasons why goal scoring rates for 3-on-3 hockey are all over the place. To get to 3-on-3 hockey in the past, teams were already a man down. So there really isn't any reliable indication of how often teams will score with their best three man units out there.