clock menu more-arrow no yes mobile

Filed under:

Analyzing the Sabres shot metrics

The numbers behind the Sabres’ shot metrics, both for and against, tell an interesting story through 48 games

NHL: Buffalo Sabres at San Jose Sharks John Hefti-USA TODAY Sports

The performance (and perceived regression) of the Buffalo Sabres’ defense had been a topic of discussion leading up the the All-Star break. A healthy contingent of the fan base has expressed the belief that the Sabres’ blueliners have performed poorly as of late, following a solid start to the season. For that reason, I took a look at the team’s shots-against metrics to try and determine if an actual regression has taken place, and if so, is it being exaggerated?

While digging myself deeper and deeper into the hockey analytics rabbit hole, individual and group performance trends began to distinguish themselves in the forward ranks as well. For a few players in particular, their shooting trends serve, in part, as an explanation for recent success (and in some cases, recent failure).

What started as an article about team defense evolved into a general shot-analysis piece where we try to make sense of current shooting trends and determine the reasoning behind them.

Individual Forward Trends

The first player that jumped off the screen was Evan Rodrigues. Dressing in 42 of the team’s 48 games so far this season, the 25-year-old has registered at least one shot in 40 of them.

For comparison sake, we’ll look at the first 20 games where he registered a shot and see if any improvement has been made over the past 20. In the first half of the 2018-19 season (to date), he registered 57 shots, converting one goal for a shooting percentage of 1.75.

Not great.

Part of his lack of success in the early-going could have to do with the distance from which he was taking shots. Through the first 20 games of the season, his average shot distance was 44.18-feet, which is one of the highest averages among Sabres forwards. His shots-on-target percentage was pretty low as a result at just 70.18.

The good news is that he’s improved significantly in that regard over the last 20 games. Not only has he registered more shots (67) in the second half, he also has four goals in that span. Fortunately, his shooting-percentage has also improved to 5.97 (still not great, but much better).

The reason for this change likely has to do with two factors. First of all, his average shot distance has been reduced by over four feet (40.11). As a result, his concentration of high-danger chances has also experienced a significant jump. The same can be said for his on-target percentage, which has gone up by seven-percent.

The chart below shows Rodrigues’ individual shot map over the first 20 games. The chart below that shows his shot map over the last 20. A much higher concentration of his shots are being attempted from the slot, as opposed to the outside where they had been coming from, almost exclusively. It also explains the steady increase he’s experiencing in his expected-goals average (which it’s important to note, has been up-and-down all season).

Casey Mittelstadt is a similar case, but perhaps not as obvious at first glance. Like Rodrigues, he has registered a shot in 40 games this season, so his split-analysis will be the first 20, versus the past 20 as well, for consistency’s sake.

Unlike Rodrigues, Mittelstadt’s shot-count has actually gone down from 49 in the first 20 games, to 43 in the last 20. The good news is, the shots he’s taking are of higher quality. In the first half of his games played, his average shot-distance of 33.99 isn’t objectively “bad” per se, but he’s managed to reduce it to an even greater degree than Rodrigues.

Over his last 20 games, Mittelstadt’s average shooting distance is 28.25. Unsurprisingly, his on-target percentage has gone up by more than five-percent as a result. Just like we saw before with Rodrigues, Mittelstadt’s high-danger shot-concentration has improved. While that trend hasn’t resulted in a significant uptick in goals for him, it should, as long as it continues.

First 20 games:

Past 20 games:

From a shot-quality improvement standpoint, players like Tage Thompson have also shown progress, just not to the same degree. One player who also stood out, but for the wrong reasons is Jason Pominville.

Now, let’s get one thing straight. Pominville’s regression here isn’t really a result of anything that he is specifically doing differently. More than likely, the disparity lies in the fact that he spent a great deal of time with Jack Eichel and Jeff Skinner at the beginning of the year.

Pominville’s season split is a little different that Rodrigues and Mittelstadt. With 34 games where he’s registered a shot, his split will start at 17 games.

The first thing that jumps out is the amount of shots he’s produced. Through the first 17 games of this sample, he averaged 3.29 shots per game. In the past 17, he’s only averaged 2.59.

To some degree, it explains his massive drop-off in goal production and expected-goals average, but again, more of that has to do with who he’s playing with. The opportunities he received alongside Eichel and Skinner resulted in an extremely high shooting percentage of 16.07 in the first half. That number has been reduced drastically, down to 2.27-percent since.

As you can see below, nearly all of his goals came from down low when he was on the first line. Since being moved down the lineup, those opportunities obviously haven’t presented themselves as often.

First 17 games:

Past 17 games:

This ties back into the lineup optimization discussion. Pominville’s respective shot-distance averages haven’t experienced a significant change. He’s always shot from further out. So again, it doesn’t really have much to do with anything he’s done differently as an individual. It has everything to do with his linemates. Based on this information, perhaps it might not be a bad idea for Phil Housley to reunite him with Eichel and Skinner and shift Sam Reinhart down to help catalyze the second line.

As a whole, the overarching issue that the Sabres have experienced as a team, is their group shot-distance average. No team in the NHL has shot from further out this season.

Buffalo is averaging more shot attempts per game this year, but perhaps a higher emphasis needs to be placed on shot quality versus quantity. At the moment they have registered the fourth-fewest high-danger chances in the league. That’s not going to cut it moving forward.

Is the Defense Really Regressing?

Short answer: yeah, a little but what’s being perceived as a severe regression is actually a combination of things.

We’ll do something similar here where we split the season into two even portions between the first 24 games and past 24. Through the first half of the games played this season, the Sabres allowed 207 high-danger chances against, which equated to around 8.63 per-game. That mark isn’t too far off from the league median.

Over the past 24 games, that number increased to 225 (roughly 9.37 per-game). This 8.69-percent uptick has resulted in an increase of high-danger goals-allowed (22 through the first 24 games in comparison to 25 allowed in the 24 games since).

On the surface, that doesn’t seem significant enough to cause concern. What’s a bit more alarming is the rate of scoring changes that the Sabres defense has conceded as of late. During the first half, they were allowing 19.79 scoring chances per game. In the second half, they’ve allowed 22.46. Again, not great, but is it enough to call their overall performance a regression?

The Sabres’ goaltenders helped cover up some of the team’s deficiencies on the blue line, which created the illusion of a vast improvement at the beginning of the year (yes, they improved but the proficiency in net made them look that much better).

As Chad DeDominicis wrote earlier in the week, Carter Hutton and Linus Ullmark have experienced a recent regression of their own. In the same way that their lights-out performances made the defense look great at the beginning of the season, their recent struggles have had a similar, converse effect.

The combination of a reduced high-danger save-percentage in the last 24 games paired with a slight uptick in high-danger chances allowed, have washed out some of the improvements we’ve seen from the forward group (from a win/loss standpoint). If the players on the back-end can get back to the level of play that they exhibited in the beginning of the season and the forwards can keep on their current trajectory, the wins will come.