The Washington Capitals may have been defeated by the New York Rangers in overtime of Game 7, but the Capitals' biggest struggles - the ones that led to that Game 7 even being necessary - occurred at the start of games.
"For all the cultural and structural improvements Trotz brought into Washington, for all the individuals enjoying career years and the old starts shining and the national analysts taking notice, the Capitals had still not fixed their slow-start woes, a compounded problem against the Rangers, one of the NHL’s fastest teams off the hop."
- Alex Prewitt, Washington Post
While the team's inability to start strong was apparent during the playoffs, it was actually an issue that caught our attention much earlier. At the time we didn't know what the cause of the Caps' first-period woes were... and honestly we still don't. But thanks to the never ending information provided by our very own Muneeb, we're at least able to look at the game in more detail and perhaps gain a better understanding of the symptoms of the team's poor starts.
Below is a graph that details how the Capitals played during five minute segments throughout the entirety of the regular season and playoffs.
It's important to remember that there is going to be some noise in this data and that these numbers are not score-adjusted. There are a lot of shot attempts throughout the entirety of a team's full campaign but we are significantly reducing the size of our sample by looking at the five minute segments independently.
With that being said, let's look at the two segments that really define slow starts, minutes 0-10:
|Game Time (Minutes)||Total 5v5 Minutes in Bin||CF%||GF%|
|Rest of Game||3720.1||52.0%||54.2%|
Is a small stretch of bad games skewing our already small-ish sample? If the problem is slow starts, why are minutes 5-10 significantly worse than minutes 0-5? While this may seem counterproductive, let's take a look at the data in even smaller segments to try and understand whether there is anything to this phenomenon.
The data here is segmented into five parts. The first four represent the regular season in quarters, while the fifth is the data from the playoffs. There is, as mentioned above, a lot of noise - but there is some general consistency in the first ten minutes, specifically in the poor quality of play reflected in minutes five through ten.
Slow starts were a leading narrative in the playoffs and the data really underscores the issue. The Capitals posted a CF% of 33% in minutes 5-10 and 40% in the first ten minutes overall in their 14 playoff games. The team was outscored six to one within that sample. Here is another way to look at the data:
"The last two games, we've played teams that have played the night before. They've jumped over us in the first period, and even into the second period.
"One final push in the third, we say, ‘Oh we tried real hard at the end,' but it's not enough. It's not acceptable for me either. " - Barry Trotz (Washington Post, 3/15/15)
The bottom line is that, while the Caps were a good team last season, they needed to do more at the start of games. If we isolate the regular season the Capitals starts do look a bit better; they controlled 48.1% of the shot attempts and had a GF% of 50, which isn't great but is still a lot better than the playoff numbers referenced above.
It should be said that this trend might not continue moving forward, especially if Trotz identifies it as a specific area to address once training camp gets underway. But with or without predictive value, this data does describe what we saw on the ice from the Capitals last season - a team that simply couldn't always be relied upon to come out of the gate strong.