The Maple Leafs are deep everywhere but right defence. I bet you’ve heard that one before. And like most things that get repeated over and over, it’s become truthy if not actually true. The signing of John Tavares nicely covered over the loss of James van Riemsdyk as far as points production goes, and he made the Leafs the second deepest team at centre in the NHL, but we may have been mistaking quantity of wingers for quality and overstating the depth there.

Quantity over quality will get you a long ways in hockey. If you need to pick just one, you want quantity of shots, but having both is better. Put the two together, that is weight the shots for quality, and you’ve got Expected Goals, and that’s what I’m using to look at the wingers down the left side on the Leafs. (Offside Review is the source.)

Starting at the top, remember when the idea that Zach Hyman was the Leafs 1LW was the scariest proposition ever?

Zach Hyman

The eye-test says he and John Tavares have clicked on the ice fairly easily, and they have some similarities in their game. Tavares is nearly as good at puck retrieval as Hyman himself, with the bonus being his passing and shooting are an order of magnitude better. What Hyman brings to the party is that he drives play. What this means is that Hyman helps create a situation where his line shoots the puck more than the other guys. He brings the quantity and Tavares layers on the quality. Or so goes the theory.

Among the forwards in the top nine, the top four in Expected Goals For percentage (xGF%) at five-on-five are all current or former members of the Tavares line. Hyman is at 51.4, and he marks the line where the forwards below him are all below 50 percent. It’s notable that Hyman and the rest of his line are worse at Corsi For percentage, so they are covering over some of their mild quantity issues with quality.

That good Expected Goals percentage is coming entirely from the offensive side of the ledger. Hyman’s line gets very high Expected Goals For per 60, and not so hot Expected Goals Against. Both the top lines are not executing defensively well, and while Hyman is not as horrible as the regulars on the Matthews line, he’s not far off.

Hyman looks good because he’s on the ice for actual goal scoring over expected and goals against well under expected. In other words, he’s been lucky to be on the ice for some of the best goaltending on offer this season so far. This is something you see often in small chunks of time on ice. The effect of the normal variance in on-ice goaltending from one skater to another is magnified by the small sample. On-ice goaltending is not the player’s doing, so Zach Hyman is not making Frederik Andersen better, but he’s definitely participating in making his job harder. He is the fourth worst forward for Expected Fenwick Save Percentage (xFSv% - which is the percentage of all unblocked shots that aren’t goals). The Matthews line are worse, and Tavares and Mitch Marner are essentially tied with Hyman.

All of that produces a Goals For percentage of 62.5 which sure makes him look like the best left wing on the ice, even if the underpinnings of his success are shaky in places.

In terms of personal point production, Hyman has zero goals, although he has an individual Expected Goals of three. He has two primary assists, and an individual Corsi For per 60 minutes (iCF60) of 11.21, which is third lowest of all forwards. This is good, not bad. Hyman’s job is not to shoot the puck. His low assist rate is hard to worry about, as normal variance over 12 games could be doing anything to points numbers. We should never expect him to personally score over his expected goals, but at some point, he’ll get near it. He finished last year with an expected goals of 17 and an actual of 13.

Patrick Marleau

If Zach Hyman at least looks like he has value, even when you dig below the illusion of his GF%, Patrick Marleau looks like he hasn’t actually started playing yet. And the numbers aren’t prettier.

Bearing in mind that he’s only been in 12 games, and he might just be having trouble adjusting to playing with Auston Matthews, Marleau has still performed extremely poorly this year. Something is really wrong with what he’s doing on the ice.

His results show a massive divergence from his recent normal performance. He has the worst xGF% of any regular top-nine forward with 47.9 , a number almost identical to his CF%. He has neither quantity nor quality. He’s getting to those dreadful percentages in exactly the same way Hyman is getting some good ones. It’s a mix of very good xGF60 and very bad xGA60. The difference comes in the scale of the badness on the xGA60. It’s really terrible. The only person with a worse number among Leafs forwards is Matthews.

This line, naturally, scores over their very high expected goals for rate, and has been getting good goaltending, but not the way over expected number the Tavares line has, so Marleau’s GF% is only 55% and might actually be more sustainable than Hyman’s.

Now let’s pull back and look at the whole league. Marleau is the fourth worst forward in the NHL this season in total on-ice Expected Goals Against (not rated per 60) at 10.68. His actual goals against is 10. The three forwards worse are Patrick Kane, Jakub Voracek and Nick Schmaltz (Kane’s linemate). So the Matthews line isn’t the only scoring line in the NHL who is killing their offensive value with horrible play in the defensive zone. Misery might love company, but Chicago suffering doesn’t actually make the Leafs less troubled by this problem.

So is it Marleau or is it Matthews? Is it the lack of Nylander, or something else? In terms of individual results, the most astonishing thing about the Leafs right now is that Marleau has the lowest iCF60 of any forward. At 10.38, he’s shooting at a defenceman’s pace, and yes, indeed, lower than Frederik Gauthier’s rate of 12.96.

Marleau has the lowest individual Expected Goals of any top-nine player, and he hasn’t scored a single goal at five-on-five. He has two primary assists.

Turning in numbers worse than Zach Hyman in key areas is not what you expect from Marleau. If Hyman is not driving play like he should be, he’s still delivering to his ability and his pay rate in overall impact. Marleau is not, and it begins with his shooting. It’s not just a question or rate, it’s location:

The heavy pattern of shooting all around the goal mouth last year — which is normal for Marleau — has been totally absent from his play this year.

For a little perspective, I just saw an analysis today that showed David Pastrnak has been shooting more, but of less quality this year. That is unlike his career norms so weird things happen at the start of the season. However, Marleau is shooting much, much less, and from less dangerous locations. The high Expected Goals For is all coming from Matthews and the right winger of the day, and at times it’s hard to tell what he’s even on the ice for.

Par Lindholm

Lindholm has played a mixed bag of fourth-line centre and third-line left wing. If I isolate out his games played on Nazem Kadri’s wing, I get seven full games, leaving off the half-game where Matthews was hurt and Lindholm moved to centre.

The difference between his results as a winger and as a centre are stark. In his winger games, he has an xGF% of 52.8 and a CF% of 55. Next time I say that strength of teammates has the biggest effect on a player’s results, I want you to remember this example.

As a member of the Kadri line, Lindholm was on the ice for four goals for, just over expectation, and one goal against, which is under by two and a bit. This gave him a happy GF% of 80, which I don’t think is sustainable, but the expected numbers look very, very good.

By the eye-test, Lindholm checks out as a forechecker the equal of Connor Brown, with a little better offensive acumen, but maybe not as good of a shot. He struggles at going from getting the puck in the defensive zone, which he’s very good at, to doing anything useful with it (also known as Leafs Disease). This might be the main reason, beyond Kadri, why his numbers as a centre are so bad. And they are.

His total 12-game set of results have him with an xGF% of 47.8, the same terrible number Marleau has, and a CF% of 51.

Individually, his shot rate is okay, not great, and it puts him in with the second-tier forwards like Kapanen, and Brown. He’s scored one goal as a winger, dead on his Expected Goals, and he also has two primary assists.

Given a chance to play more centre with wingers like Brown and whoever is on the left side long term, his results as a C might pick up. But as a winger, he’s been better, and he’s performing better than Marleau.

The other Three Left Wings

Now it gets complicated. All of Josh Leivo, Andreas Johnsson and Tyler Ennis have played some left wing, and for Leivo and Ennis, some right wing. Most of their minutes have been on the fourth line. Most of the games they’ve appeared in, they’ve played under nine minutes, and it’s really hard to take extremely low minute per game results over a very few games seriously.

In terms of individual offence, Ennis leads in Expected Goals, although half of his came from his time on the top line. Josh Leivo still hasn’t managed to string together enough shots of quality to get to one whole Expected Goal, and Johnsson is much worse at only 0.26, the lowest on the team for forwards. Which, yes indeed, means lower than Gauthier.

Johnsson has been shooting almost at the terrible rate of Marleau, while Ennis and Leivo are in Lindholm’s range of lower-tier forward.

In terms of on-ice results, their xGF60 (rated so they’ll get some allowance made for really low minutes) are not good either. Johnsson’s are very low, third worst for forwards, Leivo’s are low, but in the range for a checking line forward, and Ennis is just scraping top six territory with 2.58 — goosed a little by his time playing with Matthews.

There’s nothing about their results to say any one of them should get more ice time or can be anything but an okay third line replacement player on his best day. Ennis and Leivo have the edge over Johnsson by quite a bit. My eye-test says part of their problem is that Ennis and Leivo play together a lot and they try to do the same thing, go after the same puck, go to the same open area, and they may well get better and communicate their roles to each other better as time wears on. I’m not sure I even have a theory for how horrible Johnsson has been.

Conclusion

This all seems depressing, but it shouldn’t be. At the moment we have a swing in team CF% of 62 to 40. That will contract toward the middle, and some of the Leafs players will revert to their own norms too. When Matthews returns, his line will likely stop being the second worst in the NHL at xGA.

Zach Hyman will score a few goals, and likely have better Corsi numbers as we go along. Par Lindholm is already doing more as a winger than we should expect considering he never plays wing. The real issue with Lindholm might be that he’s too good to ever play less than nine minutes per night.  The three guys gunning for one left wing spot on the fourth line are all good at different things, and they’ll sort themselves out in time. If they take turns all season, that’s not a bad thing.

The problem is Marleau, and time might make it go away. Time on Kadri’s line might help him find his old groove. But if it doesn’t, then the door is open for one of the others to step up. I don’t think Mike Babcock is unaware of this problem, either. When Matthews exited the game hurt, Marleau lasted one try on the top power play unit. He was replaced there by Kasperi Kapanen, but the most recent practice has him back there in a new configuration.

If you aren’t thinking thoughts about Kapanen playing his off wing to stay in the top six once our missing players return, I sure am. But more than that, I don’t want there to be a top six. I want the Kadri line, once Matthews comes back, to be a legitimate part of a top nine, and he needs someone to score some goals for that to happen. So maybe the answer to the weak left wing is right there on the strong right side one way or another.