Last year, I looked at the Leafs situation on the morning of July 1, and I said that it’s good to be broke on free agent day. It is. Being broke keeps you from doing anything foolish, so if you sign a big free agent, you need to be sure it’s the right move. I think when we discuss the salary cap and discover there isn’t cap space to spend like an insomniac with Amazon one-click enabled, we assume that’s bad. But having to always make smart choices drives player evaluation and leads teams to putting better players on the ice and paying them appropriately. Most of the time.
The Leafs are likely to be broke again on free agent day this year, and they have to remove a player from the roster to add a player making more than ELC money. Kyle Dubas has to manage in the NHL in this coming season with this flat salary cap and with four top players who already earn so close to half the salary cap, we might as well just call it that.
Last year’s examination of the broke Maple Leafs on July 1 centred around the concept of finding who had the most inefficient cap hits compared to what they were producing on the ice, and suggesting they be moved out. That was easy. That was so easy, that I could have guessed all the moves (the Leafs’ half of them, not the other side because no way I would have guessed Cody Ceci) without ever looking at a single number. But the numbers showed that moving Patrick Marleau, Nikita Zaitsev, Connor Brown and yes, Nazem Kadri, made the most sense.
There was a lot of resistance to the idea of trading Kadri at the time, and a lot more now, and part of that comes from the idea that if you trade away a player, it’s because you think they’re bad, overpaid or both, and you don’t want them. That’s often true when new management is cleaning up the messes of the past, but those days are gone on the Leafs, and the inefficiencies are harder to find.
If you look for inefficient cap space this season, you will discover the expiring UFA Ceci. But expiring UFAs aren’t the point, so I looked at the players likely to actually be on the Leafs next season using nearly the same method as last year, where I compared AAV to WAR. I’m going to switch to Standings Points above Replacement this year just because that measure is a little more intuitive, but it’s all the same calculation as WAR or GAR, just expressed in different units.
XAR measures isolate an individual player’s contributions in various areas of the game using the best data available to determine what that player added to the team in a given year above what a replacement-level player would have done in the same situations. Frederik Gauthier was approximately replacement level on the Leafs this season.
Disclaimer: This is a totally unfair way to judge inherent player value. It’s one year of results, and it ignores some of the realities of how salaries are compressed at the bottom by the minimum and compressed at the top by the salary cap. It also ignores that other market forces dictate how much a player is paid. I just put in an even $1 million for the unsigned RFAs and Jason Spezza, because the players at their end of the list are the least important, so that’s close enough.
AAV vs SPAR
| Team | 2020-21 AAV or Estimate | % of Cap Ceiling | SPAR in 2019-2020 | % of total SPAR (35.10) | Difference |
|---|---|---|---|---|---|
| Mitchell Marner | $10,893,000 | 13.37% | 2.7 | 7.69% | 5.67% |
| John Tavares | $11,000,000 | 13.50% | 2.9 | 8.26% | 5.23% |
| Kasperi Kapanen | $3,200,000 | 3.93% | -0.2 | -0.57% | 4.50% |
| Denis Malgin | $1,000,000 | 1.23% | -0.4 | -1.14% | 2.37% |
| Rasmus Sandin | $894,167 | 1.10% | -0.3 | -0.85% | 1.95% |
| Andreas Johnsson | $3,400,000 | 4.17% | 0.8 | 2.28% | 1.89% |
| Morgan Rielly | $5,000,000 | 6.13% | 1.5 | 4.27% | 1.86% |
| Martin Marincin | $700,000 | 0.86% | -0.3 | -0.85% | 1.71% |
| William Nylander | $6,962,366 | 8.54% | 2.5 | 7.12% | 1.42% |
| Frédérik Gauthier | $1,000,000 | 1.23% | 0.1 | 0.28% | 0.94% |
| Jack Campbell | $1,650,000 | 2.02% | 0.4 | 1.14% | 0.88% |
| Justin Holl | $2,000,000 | 2.45% | 0.9 | 2.56% | -0.11% |
| Auston Matthews | $11,634,000 | 14.27% | 5.4 | 15.38% | -1.11% |
| Pierre Engvall | $1,250,000 | 1.53% | 1.2 | 3.42% | -1.89% |
| Alexander Kerfoot | $3,500,000 | 4.29% | 2.2 | 6.27% | -1.97% |
| Jake Muzzin | $5,625,000 | 6.90% | 3.5 | 9.97% | -3.07% |
| Zach Hyman | $2,250,000 | 2.76% | 2.2 | 6.27% | -3.51% |
| Frederik Andersen | $5,000,000 | 6.13% | 3.4 | 9.69% | -3.55% |
| Ilya Mikheyev | $1,000,000 | 1.23% | 2.1 | 5.98% | -4.76% |
| Jason Spezza | $1,000,000 | 1.23% | 2.5 | 7.12% | -5.90% |
| Travis Dermott | $1,000,000 | 1.23% | 3.5 | 9.97% | -8.74% |
Evolving Hockey’s SPAR model really likes Travis Dermott, even though the fanbase is consumed with disappointment over him. I don’t think the Leafs are disappointed, and I don’t expect them to trade him. I think he does bring value, just maybe not quite as much as you see here. Goalies get very high ratings compared to skaters.
SPAR of approximately 4 puts a player in the top 10% of all NHL seasons over the years used to build this model. Which makes them very rare. The next step down of good players in the two to four range make up only 18% of all NHL seasons. Almost 70% of player seasons hover around -2 to +2.
The difference column on the chart is percentage of cap hit minus percentage of SPAR, so negative means the person is underpaid relative to their teammates. Eagle-eyed readers will note that the total SPAR used is actually less than the total of that column. Enough of the results belonging to the players not shown were negative, and that makes the team total less than the total of this core group.
Sort it by % of the Cap and start with Auston Matthews. He looks underpaid this season, but the rest of the big four don’t. The inefficient cap hits are always expensive star players, though, because the rarity cost is part of their AAV. Last year Matthews looked overpaid, just because measures like WAR or SPAR fluctuate year to year. So while I understand why everyone will zero in on Mitch Marner, go back to just the SPAR column and ask yourself how you replace that if you trade him, because players who add what he adds are expensive and rare. His SPAR was also 16th highest in the NHL in 2018-2019.
You can see that the most expensive players, Dubas’s core, generally had the highest SPAR this season. He’s not wrong to have confidence that his core is good. Where they stand relative to the league is another story, as is how they compare to last season. But ignore them, and ignore the cheap players at the bottom, and you’ll find the names you already know are more likely to be moved out not to dump a player, but to get a return of a new type of player. They’re in the middle of the AAV list running from Kerfoot to Campbell.
Take both of them off your list of potential trade pieces. The Leafs aren’t trading their backup or their third-line centre, not unless the return includes a replacement of the same type. That was last year’s trick, and Kerfoot is working out as 70 - 75% of Kadri at five-on-five, with Kadri’s power play value taken up by other forwards like Marner.
That leaves the following:
- Andreas Johnsson: not a great deal this season (fair value on last season’s production)
- Kasperi Kapanen: very bad deal this season (good deal last season)
- Zach Hyman: fabulous bargain (good bargain last season)
- Justin Holl: fair value this season (will look even better if someone gets promoted over him)/
Trading Zach Hyman would be very, very foolish unless there was someone better looking to supplant him as a top-six winger. There isn’t, and trading the only player of his type just for the sake of asset management is an even worse idea. Hyman is paid as low as he is because his skillset is undervalued. That’s what a bargain player is in hockey. And when you sell one, usually you’re getting back a player with a highly valued skillset, and the result is called losing the trade.
You can’t trade Holl, even if he should be used a little less harshly, because he is the top right-shooting defender.
With those two choices thrown out as well, we arrived where we knew we were heading, just with a better idea about value of the two players who are most likely to be moved.
Johnsson is older, has a higher cap hit, and plays a second-unit (on the Leafs) quality of power play. Kapanen is more versatile, with some power play and very good penalty killing skills. They both look worse via recency bias than they likely really are.
Ideally, the Leafs would want to play those two with Kerfoot as the third line. It would give that line an identity, and it would mean that there are four wingers on the team better than Johnsson. Which, given the middling nature of his contributions over the last two years, is what good forward quality would look like.
In the ideal world, you could trade either one of those two for a $3.5 million AAV defenceman and put Ilya Mikheyev in to replace them on the third line, and the obvious player Mikheyev can replace is Kapanen, who seems like he’d return a slightly better quality of defender.
I think that’s exactly what the Leafs will do, and yet, the problem is that there aren’t four wingers better than Johnsson so the calculation is more complicated and likely involves playing Nick Robertson in the NHL now.
He looked better on the power play in four games in the playoffs against really tough defending than he did at five-on-five, but he didn’t look absurd on an NHL roster either. I don’t see a lot of benefit to anyone but the Peterborough Petes to sending him back to the OHL, but I also don’t want to overrate his likely contributions. He’s the future not the present.
Alexander Barabanov is a total mystery, and I’m essentially ignoring his existence because who knows if he’s Miro Aaltonen, Par Lindholm or Ilya Mikheyev.
In a really ideal world, if the Leafs had hit it big on every single winger they drafted and hadn’t needed to trade prospects to get Jake Muzzin and Jack Campbell, they could trade both Johnsson and Kapanen and seamlessly promote replacements in the system. They don’t, and that’s not Dubas’s fault, even though people will say it is. It’s not the personal fault of the previous GMs either. No one hits on every draft pick, and the Leafs have only had a professionally-run prospect development system for five years. If Dubas tried to move both of them, he’d be taking a huge risk by spreading his roster as thin as it could go at forward. A few injuries would leave the top line Pierre Engvall - Auston Matthews - Denis Malgin, and that’s something the Oilers would do.
Frederik Andersen is an interesting case because this model really does pour on the love for goalies, and it’s amusing to see Andersen’s worst season credited with this much positive impact. And yet, he did make a lot of saves, he just didn’t make enough. Given his contract expires after next season, it’s not a surprise that the same logic that was applied to Nazem Kadri last summer is being applied to Andersen this offseason.
You can’t trade Andersen one-for-one for a defenceman and walk away whistling. You could trade Andersen either directly for, or because you’ve signed, a cheaper starter, preferably younger. Then you end up with room for someone else to add to the roster. Split his cap hit into two, in other words, either all in one deal like the Kadri trade or a trade plus a UFA signing. The expansion draft throws a wrench in that plan and leads me to think the Leafs playing through Andersen’s contract and then deciding what to do is more likely.
The Kadri trade is hated because he’s scoring goals right now (what a surprise that the good player is good) and because Tyson Barrie didn’t seem to be worth it. He likely was, though. Worth the amount of cap hit he actually cost, that is. If that trade hadn’t happened, this year’s trade suggestion from everyone would be to trade Kardi for a defenceman and find a 3C who can at least do most of what he did at five-on-five, so scrap your time machine plans, that one can’t be fixed.
There is no addition by subtraction move here. There is no Zaitsev or Marleau to move. The solutions aren’t obvious, and every fix creates other problems, so the task for Dubas is to make the team better in one area while not making it too much worse elsewhere. And this is why people love rebuilds. They’re simpler, and hopes stay high.
