Author: Justin Chiodo, June 2016
Winning Percentage has always been one of the most basic and fundamental statistics in understanding a team’s regular season performance in Major League Baseball, the NFL, and the NBA.
However in hockey leagues, winning percentage has never been used because the NHL, (and all hockey leagues I know of), use a point-system to produce their standings, which always has accurately reflected winning percentage anyways.
Prior to the introduction of the OTL result (overtime losses), hockey teams win-loss records were very simple. There were only 3 possible results.
Traditional Single-Game Result Possibilities
WIN= 2 POINTS in the standings- (regardless of whether win occurred in OT or regulation)
TIE= 1 Point for a tie game
LOSS= 0 Points- (regardless of whether the loss occurred in OT or regulation)
The need to calculate winning percentage in the standings did not exist because the NHL Point System accurately quantified the on-ice results. In other words, the standings were always identical using the NHL Point-system as they would have been if the standings were theoretically arranged by winning percentage.
However, beginning with the introduction of the OTL result in 1999/00, the accurate correlation between NHL Points and winning percentage was undone. No longer would the Points also be an accurate reflection of the teams’ winning percentages.
Two new single-game results now existed that were not possible beforehand:
1. A team could now LOSE a game and still achieve 1 Point in the standings.
2. A team could now WIN a game, yet their opponent still received 1 Point in the standings.
Over the past 16 seasons, the NHL’s standings have not quantified single-game results with nearly enough accuracy.
*An Aside Regarding Shootouts…
It’s important to note that in a lot of fans’ minds and certainly in absolute fact as well, winning/ losing a game in a shootout is a very different result than winning or losing during true hockey play.
A strong argument can be made that the ‘shootout win’, should be worth less than it currently is, and could be quantified more accurately in the standings. However, I will cover this topic in a future article, and focus on the most problematic issues for now.
So beginning in 1999/00, the NHL went from having 3 possible single-game results, to now having 5 possible results.
(It’s imperative to understand that the standings- and all possible achievements like qualifying for the playoffs, winning a Division Title, winning the President’s Trophy, etc.- are all awarded based on Points.
Modern Single-Game Result Possibilities
1. REGULATION WIN
‘Team A’ gets 2 Points. ‘Team B’ gets 0 Points.
‘Team A’ receives 100% of the points awarded for the game.
2. REGULATION LOSS
‘Team A’ gets 0 Points. ‘Team B’ gets 2 Points.
‘Team A’ receives 0% of the points awarded for the game.
3. TIE - (*Ties no longer exist beginning with 2005/06 season)
‘Team A’ gets 1 Point. ‘Team B’ gets 1 Point.
‘Team A’ receives 50% of the points awarded for the game.
4. OVERTIME/ SHOOTOUT LOSS (OTL)
‘Team A’ gets 1 Point. ‘Team B’ gets 2 Points.
‘Team A’ receives 33.33% of the points awarded for the game.
5. OVERTIME/ SHOOTOUT WIN*- (has never been tracked by the NHL)
‘Team A’ gets 2 Points. ‘Team B’ gets 1 Point.
‘Team A’ receives 66.66% of the points awarded from the game.
The most urgent problem with the current format is that Regulation Wins and Overtime Wins are lumped into the same category in the teams’ win-loss records. They are both filed under ‘Wins’.
However winning a game in regulation, in which the winner receives 100% of the points awarded for that game, is a very different performance result, than a win in Overtime/Shootout in which the team only receives 66.66% of the points awarded for that game.
Another way to outline this problem, is that in a standings format where Points are paramount, its not only important to consider the points a team earns, but it’s also important to consider the points a team concedes. In summary, a win in which a team concedes 1 point to their opponent cannot be worth the same as a win where a point was not conceded.
Over the years I have had dozens of conversations with hockey fans that express frustration with the modern win-loss record format. Even at the time, I despised the change made in 1999. It’s almost impossible to accurately interpret a team's true performance from a glance at their win-loss record with the modern format.
The NHL has made an absolute mess of a team’s record. It’s basically undecipherable from a quick glance. I had to do a ton of research just to compile how many overtime/ shootouts wins each team had accumulated on a season-by-season basis. Before True Champion Sports existed, this information was not available; and yet it should be basic information given the current format.
There’s an inherent problem with having Overtime games worth more than regulation games. With some games being worth 2 points, and others being worth 3 points, it creates a situation where coaches/ players are aware that playing in overtime games is beneficial regardless of the result. In general, it would be a wise strategy for NHL teams to try to get to OT as often as possible. Here’s an example to illustrate:
Let’s say there was a hypothetical expansion team in 2015/16 who posted a record of 0-0-82. This team would have technically gone winless for the entire season, yet they would have finished tied for 19th out of 31 NHL teams.
The NHL did away with Ties for 2 main reasons- there were too many ties in their opinion and they thought that fans felt that ties were boring. However, now a situation exists where more games than ever are going to overtime. The incentive isn’t there to go for a win in the last 5 minutes of regulation when tied. The ‘boredom’ of ties has been replaced with the ‘boredom’ of such a high percentage of games going to overtime.
I’ve crunched the numbers and since 1999/00 each team plays an average of 18.8 overtime games per season. That’s 22.9% of their games. Approximately 1 in every 4 NHL games is a 3-point-game, and that’s why the point totals have become so inflated in recent years.
Example- Illustrating the Compromised Integrity of the Standings
In the 2014/15 season the Los Angeles Kings missed the playoffs. They were the League’s defending Champion, and had won 2 of the past 3 Stanley Cups; and had been to the Western Conference Finals 3 straight times. In short, they were in the midst of a dynasty.
They missed the playoffs with 95 points, while the Jets and Flames both qualified with 99 and 97 points respectively.
However, the Kings had a better winning percentage than both of these teams. If quantified properly, it clear that they achieved better results than Winnipeg and Calgary and should have been ahead of both of them in the standings.
KINGS- 40 Total Wins (3 of which occur in OT/SO) 95 points. .537 Win %
JETS- 43 Total Wins. (11 of which occur in OT/SO) 99 points. .533 Win %
FLAMES- 45 Total Wins. (13 of which occur in OT/SO) 97 points. .524 Win %
The Los Angeles Kings were deprived a chance to defend their Championship and add to their dynasty-run because the standings are not quantifying single-game results accurately enough.
Despite whatever profits and record growth the NHL reports, its pretty clear to me that hockey has fallen well behind football, basketball, and baseball in the North American sports landscape compared to where it was in the 1980s and 1990s.
Sports in general has boomed over the last 30 years- and the main reasons have to do with the introduction of the internet, smartphones, fantasy sports, real-time updates, streaming options, increased cable packages, using social media to connect with other fans, Twitter, etc.
However, when compared in popularity to the other 'Big Four' leagues in North America, over the course of my life I’ve seen hockey decline in relative popularity and mainstream relevance. In the United States, other than a few hardcore fanbases, hockey does not 'move the needle' the way it does for Canadians.
I believe its imperative that the NHL corrects the flaws that this article details, and simplifies team's win-loss records.