This Lightning season has not been fun for most fans so far. Part of that is because of the hangover from last season’s disappointing first round playoff sweep. Part of that is because the team has not gotten off to a good start in terms of results. But part of it is also how they’ve gotten to those relatively poor results.
Most of us in the fanbase are familiar with watching bad or mediocre hockey. Even including this recent run of sustained of success, this franchise has still seen more bad times than good. But this year is a particular kind of frustrating. Because most of the time, we’re not watching bad or even mediocre hockey. This team looks good, as would be expected from this lineup, often enough that it’s confounding how they can still be struggling to win games.
I’ve written about the goaltending to the point that some of you are probably getting tired of reading about it. Frankly, I’m getting tired of writing about it. But every time I try to dig into what’s going on with the team, that’s where I land. So here we are again.
One of the popular discussion points around the team is that part of the reason Andrei Vasilevskiy has struggled so much is because of poor defensive play in front of him. That idea is tough to support given that the Lightning are ninth in the NHL both in shots and expected goals allowed in all situations. If we limit to 5v5, those numbers get even better.
But to be fair (letterkenny voice), the discussion isn’t really that the Lightning are always bad defensively. The idea is more that they have a tendency to breakdown in ways that allow exceptionally dangerous shots. So in theory, it would be possible for them to allow a relatively low number of expected goals in total, but for that total to be comprised of a disproportionately high number of shots that would be difficult to save. And thus, maybe Vasilevskiy’s performance isn’t as bad as it might seem on the surface.
That’s a more difficult idea to test and it involves digging into the play by play data from the NHL at a shot level. Fortunately, Evolving Hockey provides that data to subscribers so I was able to try. I took several approaches and settled on one I think does a good job of illustrating the point. As usual, we’re going to look at pictures to try to answer this question.
The chart we’re going to use might be confusing at first so I’ll do my best to explain it. The vertical axis is cumulative goals saved above expected. The horizontal axis is the expected shooting percentage on each shot the goalie faced, which is typically referred to as shot danger. So as we move right on the horizontal axis, we’re looking at increasingly dangerous shots.
These plots might look like line charts but they’re dot charts. Each dot is one individual shot. That’s why they’re more dense to the left. Most of the shots goalies face are lower danger. As we move to the right, the dots get more sparse and I’ve filled in the gaps with a different color line to distinguish that.
To start, let’s look at the top 40 goalies in the NHL in terms of workload so far this season. They’re sorted by total goals saved above expected. Vasilevskiy ranks 38th.
We already knew that Vasilevskiy’s total number was not good. But what we want to try to understand from this visual is how he got to that total. Specifically, we want to know if his total number is bad because a disproportionate number of his goals allowed have come on dangerous shots.
This chart does not show any evidence for that idea. Most of his bad results can be attributed to poor performance on shots further to the left on the horizontal axis indicating that they were less dangerous. In fact, on more dangerous shots, Vasilevskiy has been a bit better as the trend of the dots stops tumbling as they move to the right.
Looking at other goalies can be helpful in interpreting these plots. Carter Hutton is a good example of a goalie who does well with lower danger shots but struggles as the going gets tougher. Corey Crawford is interesting in that he’s allowed lots of softies but has been stellar on more dangerous chances working his way all the way back into slightly positive territory.
Shifting our focus back to Vasilevskiy, instead of just looking at this season, we can look at his career as a whole.
Again, we don’t see his numbers deflated by a bunch of dangerous shots. Instead, his issues start well before the point where a shot would be considered particularly challenging. He also appears to let in quite a few soft goals at the very low end of the danger scale.
As one final look, we can break down that career performance into seasons. He didn’t play much in the first two years on this plot so the dots are more sparse even at the low end of the danger scale. But since 16-17, he’s been playing a starter’s workload.
This chart shows about what we would expect based on the numbers we’ve seen so far. None of these seasons look particularly strong aside from last year. Outside of his Vezina win, he’s never posted a positive goals saved above expected and we don’t see anything that sticks out to suggest that something unusual is happening in the pattern of shots he faces that would explain his performance.
So where does this leave us? Well, to be sure, we don’t see anything to suggest that the Lightning defense is the primary problem. Any time the opposition scores a goal, video study will reveal things that could have been improved defensively. But that isn’t unique to goals. That’s true of of every shot. A shot doesn’t happen unless the defense allows the offense to take one. Perfect defense would mean never allowing a shot and no team will achieve that.
The goal of defense is to limit the number and quality of shots. The Lightning have been strong in that area over the last few seasons. I’m willing to acknowledge that the public data used to build expected goals models could be failing to capture some of the information needed to measure shot quality more accurately. And with a more complete data set, maybe the Lightning defense would look a bit worse and Vasilevskiy a bit better. But I just can’t conceive of the data gap being so large that it would fully compensate for how bad his numbers have been to this point in his career.
After last season, like most fans, I wanted to believe that what we saw was Vasilevskiy reaching his true talent level and that he would sustain that. I’m finding it increasingly harder to believe that. Unless he goes on an incredible hot streak at some point this year, I think we’re going to have to consider the fact that he might be a below average and possibly even well below average NHL goaltender. Considering that he has a $9.5 million extension that kicks in next season, that would be a serious problem for the Lightning front office.
One potential mitigating factor could be the coaching of Franz Jean. The Lightning goalie coach who’s been with the team since 2010 has a history of goalies underperforming under his watch. Maybe his coaching is suppressing Vasilevskiy’s talent.
But if Vasilevskiy’s results to this point in his career are an accurate representation of his talent level, the Lightning have made a catastrophic error with that contract. That’s the harsh reality. And if that proves true, this could be the beginning of looking back on what might have been.
Notes: The data in this article does not include Thursday’s game against the Stars but if it did, the number would all be slightly worse.