Game 29 Analytical Recap: A Close Loss to Dallas and What the Numbers Reveal
A strong shot share, debated shot quality, new line combinations, and what Winnipeg’s underlying numbers say as the playoff threatens to move out of reach.
It’s been a tough couple of weeks for Winnipeg Jets fans.
Connor Hellebuyck was shut down temporarily for a light procedure, and the team has gone 3-6-1 since. That stretch now includes a 4-3 loss to the Dallas Stars.
Was this game more of the same, or were there positives to grasp onto despite another loss?
Yesterday on Twitter I said something in response to another tweet that I want to reiterate:
xGoals are not the probability of a specific shot. If they were, then with all possible variables included, a perfect xG model would simply be goals. Instead, xGoals measure the average scoring probability of all shots with similar characteristics within a defined set of parameters. They help us understand the underlying factors that drive scoring.
They are measurable and give scale.
They are also intentionally agnostic to factors like finishing talent and goaltending. The point is to isolate shot quantity and quality to better understand the underlying components of scoring.
Shots closer to the net are better on average. Shots taken from less of an angle are better on average. Shots with the shooter on their offhand side are better on average. Shots taken as rebounds are better on average.
And so on.
From this, we can determine the relative importance of different factors that help teams outscore the opposition:
Every xGoal model will differ depending on which variables it includes and how it calculates them.
MoneyPuck’s earlier model looks at unblocked shots, with angle and distance as variables. HockeyViz includes blocked shots and uses exact on-ice coordinates. Other models fall somewhere in between.
What’s interesting is that xGoal models can differ significantly in smaller samples and on individual shots. Once you get into full-season samples, they tend to align far more closely.
The Jets outshot Dallas 33-19 in shots on goal, and 64-44 in all shots.
MoneyPuck had the xGoal score around 4.4 to 3.3 for Dallas, Natural Stat Trick around 3.4 to 3.3 for Dallas (sort of), Evolving-Hockey at 4.0 to 3.4 for Dallas, while HockeyViz had it 3.2 to 2.1 in Winnipeg’s favor.
I cannot tell you which xGoal model is the most accurate for this specific game. In a single game the confidence intervals are simply too large.
I can say the Jets outshooting Dallas both overall and at 5v5 is a positive step. Shot quality is harder to pin down, but even having the argument be debatable is an improvement over many of their recent games.
So while it is not a point in the standings, there is a moral victory here.
But time is running out for moral victories.
The Jets are in a three-way tie for fourth place in the Western Conference Wild Card race and the schedule isn’t getting easier.
Scott Arniel came out with fresh lines for this game. The good news was that Jonathan Toews, who has significantly struggled this season, was shifted to the fourth line. The bad news was that Tanner Pearson, who has struggled even worse, was elevated from the fourth line.
This arrangement lasted for about half the game.
After the second TV timeout, Arniel swapped Pearson and Gabe Vilardi, likely searching for more offense by placing Vilardi with more offensively minded skaters in Cole Perfetti and Vlad Namestnikov.
Then, once the third period turned into a special teams battle, the lines went into blender mode. The old Connor Scheifele Vilardi line returned for a shift, as did the Perfetti Toews Namestnikov line and the Perfetti Namestnikov Iafallo line.
Josh Morrissey started getting double shifted with Neal Pionk and even once with Dylan Samberg.
Arniel kept searching for an offensive spark, and the Jets nearly came back, scoring three goals in the time it took Dallas to score one. Unfortunately, the Stars already had a three-goal lead.
And now to look at decision making and actions of individual players with microstats such as scoring chances, chance passes, zone exits, zone entries, forechecking, and entry denial tracked manually by moi…
Microstats Results
I am almost done my large midseason update on microstats for the team. That should come out Thursday. I do not believe last night’s game will change what I have written so far in a meaningful way.
Here are the microstats for this game:
