Tuesday, November 24, 2015

Washington Capitals: 20 Games In, Then and Now

The Washington Capitals are 20 games into the season.  It is their second-best 20 game start in club history, their 14-5-1 record eclipsed only by a 15-5-0 start in 1991-1992.  We will not bore you with the details of how that 1991-1992 season unfolded (hint: poorly*).  This is a time to compare this fine 20-game start with last year’s.  First, let’s take a look at a team-wide level as some selected metrics comparing this year to last:


It certainly looks as if the Caps are more comfortable with the systems and approaches of head coach Barry Trotz than they were at this time last season when it was all still new to them.  Improving by almost a third of a goal a game in scoring offense, whittling the goals against by more than 15 percent, and expending the goal differentials from a plus-0.15 to a plus-0.85 are sure-fire ways, at the high level of metrics, to improve the 20-game performance by eight points, year-to-year.

That 0.85 positive goal differential is especially interesting, not only because keeping score is the object of the exercise.  The 2009-2010 Capitals, the gold-standard for regular season performance in the post 2004-2005 lockout era, had a plus-1.05 goal differential.  That team did it with overwhelming offense.  This team does it with balance.

They also do it at 5-on-5, a weakness in this club at this point last season.  The 1.29 goals-for/goals-against ratio at 5-on-5 is a 43 percent improvement on last season, and the Caps rank third overall in that metric among the league’s 30 teams.

The special teams are not a lot different, looking at the special teams index (power play plus penalty killing percentages), but they got to their respective 20-game marks in different ways.  The power play is a bit less efficient than it was last season, a reduction of 4.4 percentage points from this time last season, and that is largely a product of the odd drop-off in power play production by Alex Ovechkin (four power play goals through 20 games last season, one so far in 20 games this season).  If he had the same four power play goals this season as last, all other things equal, the Caps’ power play would be almost the same 25.8 percent) as last year’s at this point (25.4 percent).

On the other hand, the penalty kill has improved in two important respects.  First, there is the efficiency aspect of it. The current version, with a 2.9 percentage point improvement over last season, is ranked fifth in the league (through Monday’s games).  The second item is in the shorthanded instances faced.  The Caps are down by eight from last season, a 12.9 percent reduction.  Not having to face as many shorthanded situation not only lowers the number of opportunities for opponents to take advantage of man advantages, it allows the Caps more time to work at 5-on-5, where they have been very effective.

What is a bit odd in the team-wide numbers is the possession statistics.  Score-adjusted Corsi-for and Fenwick-for percentages are not much different this year than last (numbers from war-on-ice.com).  Both sets of numbers are on the good side of 50 percent.   In an odd way, the lack of change here but the improvement of other numbers (goals-for-goals against ratio, for instance) might reflect a better understanding or roles and responsibilities this time around that influence the Caps’ ability to finish plays rather than just run them.

On an individual level, it is interesting to take a look at the top dozen scorers for the Caps this season who are returnees from last season (you will notice a noteworthy absence here, but we will get to that):

(click for larger image)

For this group, as a group and as individual, it is another case of what a difference a year makes.  Eight of the 12 have improved on their 20-game goal scoring performance over last season, although there are effects of differences in games played, year-to-year, among the players at the low end of the list (Wilson, Laich, Orpik, and to a lesser extent Burakovsky).  Eight players improved on their point totals (the same caveat about games applying).  The difference in plus-minus is dramatic, a 53-point improvement as a group, nine of the individuals improving on last year’s 20-game values.  Even shots on goal represents a group improvement, although one thing that is interesting is that Marcus Johansson, who adopted a “shoot-first” mentality to start last season, is off a bit on his shots on goal from last year’s 20-game start.

The possession numbers, reflected in score-adjusted Corsi-for percentages at 5-on-5 are broadly better.  Ten of the 12 players saw their numbers improve, last year to this, and in the case of Wilson and Laich there might be the effect of numbers of games played affecting the results (both missed a number of the first 20 games last season).

The one returnee not accounted for among the individuals is Evgeny Kuznetsov.  No Capitals has improved so dramatically, year-to-year, as has Kuznetsov.  Through 20 games last season, of which he appeared in 19, he was 1-5-6, plus-2, with 23 shots on goal.  This season those numbers are 7-14-21, plus-10, with 38 shots on goal.  Here is the thing, though.  His score-adjusted Corsi-for percentage is marginally lower this season (50.9) compared to last (51.6; numbers from war-on-ice.com).  You might consider this the change in quality of linemates and ice time.  Last season, he had more than 15 minutes of ice time only three times in 15 games.  This season it is 19 times in 20 games.  And, getting a lot of first line minutes with Alex Ovechkin and T.J. Oshie has not hurt.  If there is a difference among the skaters this season for the better, Kuznetsov is it.

At the other end of the ice, consider the year-to-year performance of goalie Braden Holtby:


A slow start last year has not been repeated this season.  His 12 wins is tied for the league lead (through Monday’s games).  Among goalies with 500 minutes played, Holtby is second on goals against average and tenth in save percentage.  He has allowed three or fewer goals in 15 of 16 appearances, two or fewer in 12 of them.  Compare that to last season when he allowed four or more goals in three of 15 appearances through 20 games and allowed three in just 9:34 before being pulled in a 6-5 shootout loss to the San Jose Sharks.  Perhaps more impressive is his rebound control – not shots, but rebounding after taking a loss on his ledger.  He was good after losses over the first 20 games last season: 6-1-1, 1.98, .928, but that is also seven instances of having to rebound from a loss.  This season, there have been fewer instances of having to rebound from a loss he sustained, but in those less frequent instances he has been extraordinarily stingy, a 1.25 goals against average and a .953 save percentage.

In the end…

Any way you look at it, this 20-game version of the Washington Capitals is quite an improvement over last season.  And if there is any validity to the idea that you get an idea of a team’s personality after 20 games, what Caps fans have is a very good team, but they knew that already.  What is more pleasant to consider is that they are much better than they were at a comparable stage last season, and there might be more improvement to come.


*  The Caps finished the 1991-1992 season with a 30-22-8 record in their last 60 games (it was an 80-game season), then took a 3-1 lead in games in their opening round series against the Pittsburgh Penguins, blasting the Pens by a 7-2 margin to put them on the brink of elimination.  However, the Caps then dropped the last three contests of the series, all by multi-goal margins.  It was the first time the Caps were eliminated from the playoffs by the Penguins after taking a 2-0 or 3-1 lead in games.  It would not be the last.