Analytics Perspective: Some Good, Bad, and Ugly from the LA Kings Season so Far

Kings GM Rob Blake brought up some of the more pressing numbers during Monday’s press conference in El Segundo, most notably that the Kings looked like two different teams over the first 24 games compared to their most recent 24.

Once an offensive juggernaut, outscoring all other NHL teams — a statement rarely uttered in SoCal hockey circles — quickly turned to a team that couldn’t buy a goal.

What do the underlying numbers suggest and how did they play into the recent head coaching change?

They say that math doesn’t lie. If that’s the case, then below is the cold hard truth from an analytics viewpoint that only math can provide.

Let’s take a look.

The Good – Penalty Kill

Again, Blake made reference to this while evaluating the season to date. The most consistent part of this Kings team this season has clearly been the penalty kill. And this goes far beyond just looking at their league leading 87% penalty kill percentage.

For the purpose of this discussion, we’ll use data compiled by MoneyPuck.

When down a man, the Kings have the best Goals Against Above Expected (GAAx) in the league. For those unfamiliar, there’s no need to overthink this or be intimidated by the calculation. Don’t worry, we’re going to try and explain everything in simple terms.

This statistic is the difference between ‘how many goals a team is supposed to give up’ [i.e. expected goals against or xGA] and the ‘actual goals against’ given up by the same team.

What does all that mean, some might be asking?

Basically, xGA really begins with expected goals (xG) – so let’s start there. What xG does is takes a variety of different variables, such as shot type, distance, and angle to calculate whether the given shot should result in a goal or not. It uses data from hundreds of thousands of previous shots and their results to predict whether a new shot should result in a goal or not. In other words, if a team (or a player) is outperforming their xG, they’re scoring more goals than a model would predict they would based upon the type of shots they’re taking. Conversely, if a team is underperforming on their xG, that means they should be scoring more based upon the type of shots they’re taking.

Back to xGA, this is just the sum of all opposing teams’ xG for a specific team. In other words, xGA is calculating what the Kings goals against number should be based upon how many goals the model predicts should have been scored by the opposing teams they played.

The Kings have given up 17 goals when in 4-on-5 situations thus far this season, but they’ve been expected to give up over 28 goals in that time. This means that they have given up 11 fewer goals on the penalty kill than the model expected them to.

For reference, the New York Rangers have the second-best penalty kill in the league using GAAx. They’ve allowed around 8 goals less than expected.

If there is one thing that does not need changing in LA, it’s the penalty kill.

The Bad – Goalie Rotations

Before the calendar flipped to 2024, Cam Talbot had been the best Kings goalie this season. With a 14-7-3 record, he posted a .924 save percentage (SV%), earning him an All-Star selection. However, after January 1, Talbot is 0-6-2 with an .873 SV%.

Of course, Talbot isn’t singlehandedly at fault for the recent struggles. Yet, he hasn’t looked anywhere near as sharp as he did to start the 2023-24 campaign. And math supports the eye test in this case.

When looking at more advanced stats like goals saved above expected (GSAx), models suggest the Kings should be playing David Rittich as much or more than Talbot for the time being. GSAx is calculated the same way as GAAx: it’s simply xGA – GA, but is applied to individual goalies rather than the team as a whole. Rittich has 7.4 GSAx in 11 games, while Talbot has 2.5 GSAx in 32 games. Breaking this down to a per-game (60-minute) basis, Rittich is saving about 0.7 more goals than expected per game while Talbot is only saving 0.08 more per game. Said differently, Rittich has been giving the Kings nearly one extra save per game.

In the spirit of fairness, there are other factors to consider when evaluating a team or player, and often those other elements can’t be considered when looking at analytics. For example, Rittich has played fewer games this season (11 games played vs. Talbot’s 32 GP). Looking at ‘per 60-minutes played’ attempts to equalize things, but it’s still not perfect. The larger the sample size, like career numbers, can also be misleading because of other factors; like, who is in the lineup, who is out injured, etc. The team around a goaltender is going to impact his numbers, for better or worse.

The question in this case is, are the numbers for Rittich sustainable moving forward?

Another question that can’t be measured is how much has fatigue played into Talbot’s drop off? With some recent rest, will he bounce back?

At least for now, analytics would suggest new coach Jim Hiller should look to start Rittich more often. Beyond the numbers above, Rittich has held his own in recent games against better teams like the Hurricanes and Rangers, and was the starter for the Kings only three wins in January.

The Ugly – Stars Underperforming

It’s no secret the Kings have faced significant challenges this season, particularly in the performance of several star players. PL Dubois, acquired during a sign-and-trade this past offseason, is currently on pace for 34 points — which would be a career-low for a full 82-game season.

On a very similar deal to 25-year-old Dubois (who has a contract carrying an $8.5 million AAV) is 27-year-old Roope Hintz of the Dallas Stars ($8.45M AAV). For comparison purposes, let’s look at some analytics on the two players.

Hintz is currently on a point-per-game pace compared to 0.42 points-per-game for Dubois, so he’s roughly giving the Stars twice as much production. What’s easy to point out is the fact Dubois and Hintz play very different roles on their respective teams, with one being a top-line center and the other being a third-line center. This is where the analytics can help even things out to a more objective view of what’s going on. Being on different lines and playing with ‘better wingers’ will account for some of the difference in total scoring, so this is why looking at goals above expected (GAx) is perhaps more important.

A player’s xG only cares about the shots that specific player has taken and how likely those are to result in a goal. Thus, GAx will provide insight into if a player is under or overperforming in their given situation.

Dubois has 3.7 goals less than expected so far this season, which is 27% worse than expected. In other words, the shot selection (according to the models) say he should have scored nearly four more goals this season.

In contrast, Roope Hintz has 7.2 goals above expected, 49% above expected, meaning he is overperforming. On the surface, Roope’s performance isn’t sustainable, while Dubois should see an increase in performance as these numbers naturally come back to what’s expected, based on their history.

For more context and trends, in Dubois’ first four NHL seasons, he had a positive Goals Above Expected number. The past three seasons have been negative, with one of those years being greater than 20% less than expected.

Digging deeper, according to Natural Stat Trick, for a good chunk of the seasons Dubois had the second highest offensive zone starts among all Kings players this season (behind Jaret Anderson-Dolan). This means he literally started his shifts closer to the opponent’s net than virtually anyone other than JAD. Several other Kings players have since caught up, yet Dubois is still hovering around 60%, which is still better than any of the Kings top-6 forwards.

Back to Expected Goals, fellow Kings forwards Phil Danault and Kevin Fiala have scored 1.4 and 2.7 goals less than expected, respectively.

Adding in Dubois, these three players are making over a quarter of the current salary on the current active roster, and are severely underperforming, offensively speaking.

In order to turn the Kings regular season fortunes around in the 33 games remaining, Hiller will need to find a way to get the most out of the trio. Resurrecting LA’s stagnant offense will be critical to improving their playoff hopes.

RELATED CONTENT:

Rob Blake on Why He Term’d McLellan and Why He Thinks Hiller is the Right Guy

Follow @mayorNHL

Lead photo via ALoImages

Note to webmasters/reporters: When recapping news or interviews from this site please remember to include a link to www.MayorsManor.com