In the Cockpit – An Analytical Dive Into The Winnipeg Jets Season (So Far..)

Welcome to In the Cockpit! This is a new quarterly segment where we examine some of the numbers behind the game. While we have taken looks at advanced stats before, this will provide a quarterly look to see how the team is trending. While wins and losses are the most important stat, it’s also important to know what goes into those wins and losses and what the future might hold for the team.

This quarterly segment will primarily be based on team stats, but there will be various other mentions of individual players or specific line combinations/pairings.

For those of you who like numbers, this is for you. For those of you who like graphs, this is for you. If you like old-school hockey and focus on goalie wins or an individual player’s +/-, then hopefully you can learn something new along the way.

And before you continue read this old but good article on why plus minus is the worst statistic in hockey.

Before we get into the actual stats involving Winnipeg, we need to lay a little groundwork. The first question regarding advanced stats is ‘why should we care?’ Honestly, it’s a fair question, especially when some advanced stats seem far too advanced for the average person to understand.

Basically, the reason why people care about numbers beyond the boxscore is because studies have shown that there are better ways to predict future success than by just looking at goal totals.

It might seem counter-intuitive at first, but it’s due to the nature of hockey. With so much going on in an average hockey game and only a few goals being scored per game, goals are awfully hard to predict just by looking at the boxscores. This is why we see teams score anywhere from 0-9 goals on any given night against any type of opposition. One game everything goes in and the next night it’s the opposite and the goalie’s stopping everything.

With such a tough to predict sport, people have been using alternate stats for years to help predict future success. That’s why the rise of corsi was such an important movement in the hockey community. Essentially, people started to realize that getting more shots than the opposition was a good thing (obviously), but also getting more shot attempts was also a good thing, even if those shot attempts were blocked or missed the net.

It’s common sense really, if one team attempts twice as many shots as the other team, they will likely come out on top. Not always, but in the long run things will probably work out well. In recent years we have seen the rise of expected goals (xG) as well which puts a bunch of numbers in a blender such as location of the shot, the type of shot (tip, slap, wrist), game situation (rebound, cross-seam pass), and shooter talent to come up with the probability that any given shot will result in a goal.

Most advanced stats websites nowadays are using a variation of expected goals for much of their analysis.

With me so far?

For those of you who have been following advanced stats for a long time, there was likely nothing new in what I just talked about. However, it’s always good to go back to the basics and remember some of the key attributes for the new way of thinking in the NHL.

Now it’s time to get into the actual stats and graphs. For the first In the Cockpit segment, I will do my best to explain what exactly the numbers mean, what the acronyms represent and what all of the charts actually mean.

To start things off, we will take a look at the most common aspect of the game, when the teams are even strength. Let’s see some numbers with the Jets at 5 on 5.

One of the most basic tests is looking at a team’s corsi for percentage (CF%). This essentially means what percentage of the total shot attempts in a game are taken by a specific team. For example, if a game has 100 shot attempts combined between both teams and team A had 60 of those attempts, then team A has a CF% of 60%. As you can tell, the obvious benchmark is 50%. If a team is above 50% that means they out-shoot their opposition. If their CF% is below 50%, it means the team tends to get out-shot. The important distinction of corsi is that it counts all shots attempted by a team. That means shots that miss the net or get blocked are factored in as well.

In terms of CF% this season, the Jets currently sit 20th at 49.43%. They are slightly under the 50% benchmark, but near the middle of the pack. The raw shot attempts suggest that Winnipeg is merely an average team at 5 on 5.

Another important takeaway from these first set of numbers it the far right column. This is the actual goals for % (GF%) It works the same way as shot attempts. If there are 10 goals scored at 5 on 5 during a game and team A has 4 of them, then their GF% is 40%.

As you can see by the chart, the Jets GF% is very similar to their CF% as both are near the 40% mark. When looking at some other teams, there are a few large discrepancies. San Jose is 11th in CF% with 51.29% but they are getting badly outscored as their GF% is only 39.81%. This is where the randomness of goals really starts to show. Theoretically, teams should be scoring roughly at the same rate they are attempting shots. We can see that isn’t the case, likely because it’s still early in the season and a few bad goals or lucky bounces can have a very large impact on goals for or against. For example, if the Jets were to allow three more goals while not scoring any, their GF% would drop to 46.43%. Because goals are quite rare compared to shots, each goal has a large impact on the end numbers.

Takeaway from chart one is that Winnipeg is getting out-shot, but only by a very slight margin.

Now on to chart two. This one features the next step in evaluating overall team play. This chart includes expected goals for percentage (xGF%) along with scoring chances for percentage (SCF%) and high danger chances for percentage (HDCF%). In case you haven’t caught on yet, all of the percentage stats are calculated the same way. In this case, the only tough one to understand is expected goals. Different models calculated it differently, but the idea is that a computer takes a bunch of inputs about a shot (where it’s coming from, who’s shooting, rebound or tip or wrist shot) and it calculates the probability that the shot will go in.

In the same manner the percentage for is a given teams portion based on the entire expected goals for and against. If Winnipeg plays Nashville and the Jets end the game with 2 expected goals and Nashville ends the game with 1 expected goal then Winnipeg would have an xGF% of 66.67%. They are expected to score two third of the goals in the game.

In this chart, we start to see some troubling trends for Winnipeg. We noticed in the first chart that Winnipeg is basically even at shot attempts, but looking at this one they are second last in the league in terms of xGF% with 43.8%. So Winnipeg attempts 49% of the shots, but they are only expected to score 44% of the goals.

Something seems off in that regard. The most likely explanation is that Winnipeg gets lots of shots, but not a lot of really good scoring chances. The other option is that Winnipeg allows an average number of shots, but those shots are exceptionally dangerous. That leads to the next column. As you can see in column J of the spreadsheet, the Jets fare much better in the scoring chance department as they get roughly have of the chances when on the ice. So why is the xGF% so low?

If you look at the furthest right column, things start to become clear. These are high danger chances. The shots that are the most likely to result in goals. Think of breakaways, rebounds, odd-man rushes, that type of thing. In terms of HDCF%, the Jets are downright terrible as they only get 39.53% of the chances. This number is easily last place in the entire NHL and it provides a very clear answer to why the Jets get an average number of shots but have a much lower xGF%. Well, not exactly clear. Things become clearer when looking at the inputs into the HDCF%. As you can see, column K and L have the raw high danger chances for and against. When you compare them to the other teams, the Jets are on the higher end in chances given up with 205. Inversely, the Jets are last in the league in terms of high danger chances on offense with only 134. So the Jets give up more dangerous chances on average and they generate the fewest high danger chances in the league. Combine those two numbers and you get Winnipeg as the worst team in the league in terms of HDCF%.

The takeaway from this chart is that Winnipeg doesn’t look good when digging into the quality of their chances. The quantity is right around league average, but the quality is the worst in the league. They give up far more dangerous chances than they generate. Breaking it down further, the Jets are poor on offense and defense as they are near the bottom of both categories. We will focus a bit more on offense as we continue.

Perhaps the best way to illustrate this is through a nice looking graph.

For all of you visual learners, you can see the Jets are the furthest left on the graph with the Rangers. That means they have the lowest xGF%. But even though Winnipeg is the furthest left, they are near the centre from top to bottom, which measures their CF%. As mentioned above, they are near the middle of the pack in terms of CF% but near the bottom in terms of xGF%.

We now know that Winnipeg is generating low amounts of expected goals. The next question is why? The Jets are skilled, fast, and have plenty of firepower up front. They should be generating tons of dangerous opportunities. One of the best ways to understand the shot quality is to look at where the actual shots are coming from.

This shot map indicates where Winnipeg is shooting from. All areas in red are where they get more shots than league average, while the areas in blue are where they get fewer shots than league average. The opposing team’s net is at the top, and the blue line is just above the 1105 minutes box.

This chart is very telling and it further supports the idea that Winnipeg is struggling to get dangerous scoring chances. The Jets are well below league average when taking shots from the slot, so it’s no real surprise they don’t get many dangerous chances. The overall effect of their shot locations reveal a -15% threat compared to the league average.

This chart is also backed up by the eye test. The Jets don’t often get into the slot and score very few goals off rebounds and scrambles in the crease. They have much difficulty getting those dangerous chances and aren’t willing to get to the dirty areas to generate those opportunities.

If you are already an advanced stats lover, this post likely didn’t contain any new information for you. As we move along, there will certainly be other issues and stats to discuss, but for now, everyone’s a little bit more knowledgeable about Winnipeg’s shot attempts, xGF% and why there’s a large variation between the two of them.

In future posts, we will examine different line combinations, special teams, defensive structure, or a few other micro-stats as they relate to individual players. For now, enjoy the Jets winning hockey games.