On Thursday night, two third period goals scored in quick succession proved to be all that the Washington Capitals needed to defeat the Vegas Golden Knights. In doing so, they became champions, and the core built around Alex Ovechkin finally earned the right to lift the Cup after years of bitter playoff disappointment.
At some point in the Cup Final, I recall reading that both Braden Holtby and Marc-Andre Fleury played relatively few regular season games compared to most starting goalies. I looked it up, and it’s true. Holtby ranked 18th among goalies in TOI this past season, while Fleury came in at 25th.
The two goalies who made it furthest in the 2018 playoffs had a relatively light regular season workload. Could this be more than coincidence? Could a lighter workload directly translate into improved playoff performance? My first thought on the matter was that a goalie who played fewer regular season games would experience less fatigue, and so would be better suited for a long and grueling playoff run. Intuitively, this theory is pleasantly logical, but does it hold any merit?
The data
To tackle this question systematically, I examined the number of regular season games played by starting goalies of all playoff teams dating back to the 2007-08 season. I defined a playoff run’s starting goalie as the goalie who played the most minutes for that team in that playoff run. I grouped the goalies by the number of series that their teams won, thus separating goalies by degree of playoff success. I then looked at the number of regular season games played by the goalies in each group.
Cup-winning goalies tend to play 7-9 fewer regular season games
The numbers in the coloured boxes show the median GP value for all starting goalies whose teams won the number of playoff series found on the horizontal axis. It’s worth noting that I prorated games played for the lockout-shortened 2013 season as if it were a standard 82 game season.
Interestingly, when we group by degree of playoff success, we can see that the goalies who went on to win the Stanley Cup generally played fewer regular seasons games than did the goalies who went on to be eliminated at one point or another. This certainly supports the hypothesis that having your starter play fewer games would help your chances in the playoffs. Let’s take a closer look at these Cup-winning goalies.
Regular season workload of Cup-winning goalies
Of these 11 goalies, only 2 appeared in 60 or more regular season games: Jonathan Quick’s 69 games in 2011-12, and Marc-Andre Fleury’s 62 games in 2008-09. Comparatively, this rate of 2/11 is quite low:
Cup-winning goalies reach 60+ GP less frequently than any other group
Only 18.2% of Cup-winning goalies reached 60+ GP, while 47.2% of all playoff starters reached the same threshold. The difference between the two figures is stark, but let’s remember that sample size is a crucial piece of context. Due to the nature of awarding a title, we can only glean a single data point per season. As such, we have just 11 data points, and that’s including 2017-18 Braden Holtby.
We can’t ignore the possibility that Group 4’s low rate of 18.2% was caused by chance. If we were to simulate 11 random trials that each independently had a 47.2% chance of producing a certain outcome, as we established is league average for hitting 60 GP, the binomial distribution tells us that there’s a 4.8% chance that 2 or fewer of the trials would produce the desired outcome. In other words, there’s a 4.8% chance that the observed statistical phenomenon can be completely explained by random chance.
Shifting perspective, this also means that there’s a 95.2% chance that the result is not entirely attributable to chance, and there’s that at least some form of relationship that exists between a goalie’s workload and their likelihood of winning a Stanley Cup. The results, though produced in a small sample size, certainly suggest that a goalie being well-rested contributes to their ability to lead their team to a championship.
So I should rest my goalie, but when?
This was my follow-up question. Accepting that a well-rested goalie is an ingredient in the Stanley Cup recipe, does it matter when that rest happens during the season?
To highlight patterns in the workload of the same 11 Cup-winning goalies, I split each of their regular seasons into thirds (Games 1-27, 28-55, and 56-82) using schedule data from https://www.hockey-reference.com. For each section of games, I examined the starter’s proportion of their team’s total goaltending minutes. For example, in Games 1-27 of Washington’s 2017-18 season, Holtby played 1162:34, which was 71.5% of all TOI for Washington goalies. The chart below shows data for all goalies, including a group median.
Cup-winning goalies tend to have their lightest workload in the season's middle third
Cup-winning starters tend to play a larger proportion of their team’s minutes during the first third (Games 1-27) and the last third (Games 56-82) of the regular season. Comparatively, during Games 28-55, they tend to play about 7% less frequently. The emphasis on the beginning and end of the season is logical: a team must win games early to build a comfortable position in the standings, and a team must win games late to enter the playoffs firing on all cylinders.
This chart suggests that the best time to rest a starting goalie is during the middle third of the season. This is not an inflexible rule, however, as we can see that there are many ways to structure rest over the course of a season and experience playoff success. Holtby, for what it’s worth, was at his busiest during the middle third of this past season and was still able to remain sharp throughout the playoffs.
Conclusion and takeaways
Over the last decade, we’ve seen well-rested goalies lift the Stanley Cup more often than not. The empirical data support the notion that resting starters more frequently, particularly in the middle third of the season, will increase the likelihood of playoff success. This means that NHL coaching staffs with championship aspirations could gain an advantage by proactively managing their starter’s workload throughout the season.
Over-reliance on a starting goalie induces fatigue and invites the risk of said goalie being unable to maintain their performance over a two-month playoff run. While teams with strong starting goalies have tendencies to lean on them heavily throughout the regular season, this may be detrimental to championship aspirations. If a coach truly wanted to maximize their team’s Stanley Cup chances, they must ensure that their starting goalie is rested enough to maintain physical and mental focus over an extended playoff run. If this can be done, the team will be one step closer to hockey’s ultimate prize.
All data taken from Natural Stat Trick (https://www.naturalstattrick.com/) unless otherwise specified
Cover photo credited to John Locher — Associated Press