clock menu more-arrow no yes mobile

Filed under:

Sloppy Firsts: Not a Habit of Highly Effective Hockey Teams

How slow starts tend to lead to fast exits

Geoff Burke-USA TODAY Sports

A week ago, that was sorta funny.

Five games later, after being outscored 7-2 in the first period, with those two coming against the League's worst team in the only game of the set in which the Caps exited the opening frame without trailing? Not so much.

Here's Justin Williamsfollowing yet another one of his team's "piss poor" starts (his words, though who's going to argue?), this one in Los Angeles:

It’s been our Achilles heel. Our starts have been terrible, can’t come back every night, that’s certainly not going to happen. So we’ve got to jump on that, that’s been a problem for us.

You know the numbers. The Caps have surrendered the first goal in 12 of their last 14 games, and simply haven't been good in the first period over the course of the season, both recently and in aggregate. That they've been able to climb out of the holes they've dug at a remarkable rate is great for their position in the standings, of course, and for their confidence as well... but maybe, just maybe, it's also reinforcing bad habits that at this point have to be considered worrisome.

How so? Consider that of the 240 individual team seasons since the start of the 2007-08 campaign, not one team that had a worse first-period goal differential than this year's Caps have right now has has won the Cup. Or made the Finals. And only two of the 77 teams with that distinction (of a worse goal differential) even made the Conference Finals.

Now that is, of course, in part because those teams - the "worse first-period goal differential than this year's Caps" teams - tend not to make the playoffs at all. Of the 240 seasons we're looking at, only 77 (32%) had a worse differential than the Caps, and only 17 of those clubs (22%) made the post-season. Eight flamed out in Round 1, seven in Round 2 and two lost in the Conference Finals (the 2007-08 Flyers and 2009-10 Canadiens). That's probably because teams that post bad numbers tend to be bad teams, not good teams that are snoozing through the first twenty (like this year's Caps).

But all-situation goal differential can be a bit wonky sometimes, so what about score-adjusted five-on-five Corsi-For percentage (i.e. the share of shot attempts a team is taking, adjusted for score effects)? They tell a similar story - less than a quarter of the teams that were worse by this metric than the Caps currently are made the playoffs, and they had roughly the same amount of success, with two notable exceptions (more on that in a moment). Take a look:


What you see along the x-axis is the teams' regular season score-adjusted five-on-five CF%; on the y-axis, playoff success ("0" means they missed the playoffs, "1" means they lost in the first round, and so on up to "5" which are the eight clubs that won it all), and the vertical red line shows where the Caps are at present.

The only team that won a Cup with a worse SACF% than the Caps have right now (48.5%) was the 2008-09 Penguins, who came in at 48.4%. Of course, they made a coaching change that season, and in the 25 games under then new-coach Dan Bylsma, Pittsburgh posted a first period SACF% of 54.4. Interestingly enough, that outlier to the left of the current Caps on the "4" line for Stanley Cup runners-up is the 2007-08 Pens.

So basically, one team (in consecutive years) has ever made it to the Conference Finals with either a worse first period 5v5 SACF% OR goal differential than these Caps.

To be sure, this post is about correlation and not causation, and it's about the "what," not the "why" (which we've theorized about before). But slow starts hampered the Caps during the regular season and into the playoffs last year, and it's a problem that's cropped up again this year (though under obviously different circumstances). If you're waiting for them to fix the problem and hoping that they can flip the switch when it matters... it's beginning to look a lot like it matters.

All stats in this post via or box scores.