In a very general sense, the top four defenders on the Toronto Maple Leafs, when they’re all healthy and in the lineup, have played very similar usage. They play with the top forward lines, they play against tough competition, and they each take about the same number of turns with the fourth line. There’s some more subtle differences in terms of score state or how the duties with the top three lines are shared out, but it’s still much more similar than it is different.

And yet one of these four is not like the others when it comes to stats, in particular Corsi Against. Nikita Zaitsev is almost as bad as Roman Polak at shots against.

Leafs Top Four Defenders as of January 30, 2018

PlayerGPTOITOI/GPCF/60CA/60CF%
Morgan Rielly47790.0216.8163.5858.3752.13
Ron Hainsey51863.3316.9358.7359.9649.48
Jake Gardiner51953.3818.6958.0559.6149.34
Nikita Zaitsev34606.7217.8456.1263.7546.82

Two things stand out. Zaitsev’s Corsi Against per 60 minutes is very different from Jake Gardiner’s. And Morgan Rielly has much better Corsi For per 60 minutes than anyone else in this group.

Why are Nikita Zaitsev’s shots against so bad?

While it’s true that his shots for (CF/60) is a bit lower than the other three defenders, the big difference in his numbers and the rest of the top four is his shots against (CA/60).

To begin to parse out why that is, let’s look at Gardiner’s and Zaitsev’s time together and time apart.

Zaitsev and Gardiner WOWY as of January 30, 2018

Player 1Player 2TOICF/60CA/60CF%
Nikita ZaitsevJake Gardiner493.2856.4760.0348.47
Nikita Zaitsevw/o Jake Gardiner113.4354.5879.940.58
w/o Nikita ZaitsevJake Gardiner460.1059.7459.1650.24
w/o Nikita Zaitsevw/o Jake Gardiner1371.2561.4157.8951.47

Mostly due to Zaitsev’s injury, Gardiner has more time away from his regular partner than Zaitsev does. But the interesting thing is that when they are together, their shots against (CA/60) is just a touch over team average at 60. Gardiner away from Zaitsev is insignificantly better at 59 and gets his improved percentage from a higher rate of shots for.  Zaitsev in his 113 minutes away from Gardiner is at 80 CA/60! That’s really bad.

We’ve joked before about the “Gardiner effect”, and it’s not really a joke. He has had a consistent positive impact on his partner’s shot rates for years. But this has to be more complex than just that.

One of the things about WOWYs that make them dangerous as an analysis tool is the temptation to infer an angel and a demon out of the two players. If you look at that chart above and genuinely think that Gardiner is good and Zaitsev is nothing but a drag on him, then you are making two mistakes. You’re ignoring that Gardiner’s considerable number of minutes away from Zaitsev is not dramatically different in results, and you’re imagining that there are no other factors influencing the difference in Zaitsev’s results.

What is happening when Zaitsev and Gardiner are apart?

Zaitsev has some minutes with most of the rest of the defenders — a lot of it is little scraps here and there made up of accumulated partial shifts together. It adds up to a small chunk of time not well spent.

Zaitsev has just over two minutes with Calle Rosen, in which they had bad shots against, but Zaitsev has never played with Martin Marincin. He has just over four minutes with Roman Polak, in which they had bad shots against; however, they had such fantastic shots for, they end up with a 53% Corsi. None of that matters to a hockey season much, but it doesn’t help Zaitsev’s overall CA/60 either.

Zaitsev played 41 minutes with Andreas Borgman and had a slightly bad CA/60 of 66. They had a low CF/60 of only 56, so their percentage wasn’t that great.  But that 66 isn’t enough to swamp out all those better minutes with Gardiner.

The main culprit here is 50 minutes with Ron Hainsey. Their CA/60 is 89 and their CF/60 is only 48, giving them a CF% of 35.  That’s enough minutes to add up to three games together, and that’s just flat out terrible. It’s that chunk of time that moves the needle on Zaitsev and makes him look worse than his usual partner does overall.

A deeper dive into those 50 minutes reveals some interesting forward lines enjoying failure together with Zaitsev-Hainsey. They saw a bit of fourth line time, but it was particularly bad fourth line time, including 4 minutes with Matt Martin and Dominic Moore that featured 15 CA and 0 CF. That’s a stunning shot rate against in so few minutes.

Zaitsev-Hainsey have featured with the Nazem Kadri line for almost 20 of their 50 minutes. The picture there isn’t pretty, and it is time spent behind Kadri’s line all season that has made the Gardiner-Zaitsev numbers as bad as they are. Get Gardiner-Zaitsev away from that sometimes failing matchup line and they soar. Swap out Gardiner for Hainsey with Kadri’s line, and yikes!

Zaitsev-Hainsey have also posted 12 unsuccessful minutes with the Auston Matthews line. A bit of time with Zach Hyman and Leo Komarov together indicates they’ve gotten some of the horrible post-power-play shifts of doom that feature all the “defensive” players on the Leafs circling fruitlessly in front of Frederik Andersen and watching the puck get worked by the other team’s rested top line.

Gardiner, meanwhile has 157 minutes with Connor Carrick and they put up such sweet numbers, you’d wonder why they aren’t the top pairing. Of course, they were getting that with 62% of their faceoffs in the offensive zone and about one third of their time with Tyler Bozak (and Mitch Marner) just whipping the puck around in the offensive zone and laughing.

Gardiner also has 104 minutes with Hainsey, and the result isn’t as bad as Zaitsev-Hainsey, but it’s not good. Their CA/60 is 65 and their CF/60 is only 54. Without these minutes, Gardiner’s stats would improve as well.

What I’m left with is this picture: Gardiner and Zaitsev are a hard-working pair.  They take the tough competition, got stuck with the Bozak line disproportionately when they were failing offensively, play with the Matthews line a lot when they are matched up against a tough line, and they do well enough at all of this to get results around team average.

However, when they split up for a little vacation from each other, Zaitsev goes and climbs Everest while being swarmed by bees and dragging a 500 lb pack, and Gardiner sits by the pool and orders drinks with little umbrellas in them.

That pool-side time with younger defenders like Carrick and Borgman is important. The Gardiner effect is so powerful, it convinced many people Frank Corrado was an NHL defender, and it’s really good for development. It’s also really useful to game outcomes to put offensively gifted defenders behind the Bozak line in the offensive zone. Goals are nice, and we should have more of those.

Meanwhile climbing the tough hills, bees optional, is Zaitsev’s job. He’s supposed to be good at this, playing the tougher minutes and handling things defensively.  And when he does that with Gardiner, life is okay. It’s just when he leaves Gardiner by the pool that things get hinky.

Look, it has to be said, none of the Leafs’ top four are getting fantastic defensive results. Many players on the team are playing just a bit over their heads, as are all of the top four defenders to one degree or another. If you want to look at Zaitsev, who I think has been tossed off the bow of a boat out at sea and told to swim for home, and see failure in his numbers, while Carrick’s laps in the hotel pool are success, well okay. But I think the reality is more complex than that.

One part of that reality is that Zaitsev’s job is unlikely to get much harder and may even get easier if the Leafs can upgrade Hainsey just a little.  But Carrick will have to graduate to a bigger pool if he’s going to stick around. There’s always a new, young prospect who can do those laps just as well.

If the practice pairings hold for today’s game, then Carrick will be on the third pair (the splash pad in this tortured analogy), and Zaitsev will be by the pool watching to make sure Travis Dermott doesn’t drown. Neither one of them will be taxed too hard by that. When Morgan Rielly is back and things return to normal, then it will be time to look again at who is handling their job well enough and who isn’t.

Now about that really high CF/60 of Rielly’s, where’d that come from …

All data is from Natural Stat Trick and is Score and Venue Adjusted.