Matt Cane has been working on a salary predictor model for free agents for a few years now, and it’s proven to be fairly accurate in years past. This year’s predictions are now available:
Here we go! Final (assuming I don't find any errors) contract predictions for the 2018 Free Agent Class:https://t.co/CwF6kTUfdZ— Matt Cane (@Cane_Matt) May 21, 2018
And as free agents are signed, he will fill in the actual results, so you can see how the model performs. This is a model designed to predict what a free agent will get, not what they “deserve”. So if you think the amounts are absurd, well, isn’t absurdity a hallmark of July 1?
This list includes RFAs as well as UFAs, so we’ll look at both.
Maple Leafs Current UFAs
James van Riemsdyk
James van Riemsdyk comes it at 15th on Matt Cane’s list, which is sorted by the Predicted AAV the model has returned for each player.
No matter how many Leafs fans, who have grown too familiar with van Riemsdyk to see him clearly, claim he is nothing but a power play tip artist, he’s going to get paid more this summer than he makes now.
The model predicts three years at $5,354,777. And let’s talk contract length. The spreadsheet shows the AAV for the length of deal that the model gave the highest probability to. In a lot of cases the percentages are fairly even across multiple terms, and the model doesn’t seem to deal with the realities of contract lengths very well. We can have the argument about how teams should give UFAs coming up on 30 shorter deals, but that requires something in the ballpark of collusion to achieve, so it’s not happening.
The model’s AAV for a seven-year deal is $6,062,262. And a six by seven is exactly what I expect van Riemsdyk to get. Given the relative shortage of scoring winger UFAs, he might get more.
Back when the Leafs signed Patrick Marleau, I said he was a shorter-term temp designed to bridge the gap between van Riemsdyk and whoever can be developed or acquired to fill his production. I think that’s still true. The cap space for a potential van Riemsdyk extension is already spent, and the Leafs will take a pass on renewing this deal.
Tyler Bozak, who at 32 is a slightly different sort of UFA to van Riemsdyk, is the next highest on the list several slots down. He might not be able to command a seven-year deal. The model landed on a three-year deal for him at $4,378,618, and the longer deals for him ramp up to around the five million mark. Some number of years more than three at about $4.5 - $5 million seems very likely for Bozak.
That’s not an outrageous price to pay for a centre, but we all know that to put Bozak in a position to succeed, for which you get a positive Expected Goal differential at even strength, you have to overload at least one and possibly two other lines. In an ideal world, the Leafs want those days behind them, but you don’t always get to live in paradise.
At the moment, the Leafs have two NHL centres. Period. They have a potential one in William Nylander, and a useful substitute in Patrick Marleau, along with some maybes in Miro Aaltonen and Par Lindholm. It’s not out of line to look at that and say that Bozak on a short deal is worth it. Turn the question around, though, and ask why Bozak would take a short deal if he gets offers elsewhere with more security built in.
Leo Komarov is next, and he is an interesting case. He was used in one way for a lot of the regular season — over his head, most would agree — and then he was used on the fourth line where he was fine. Injury kept him out of the playoffs, and Mike Babcock chose to play Andreas Johnsson over Komarov when he was healthy.
The model has Komarov at $2,856,279 for a three-year deal. This is one where I agree with the model’s projected term. The deal that is expiring for Komarov was four years at $2.950 million AAV. And while it might seem better to sign a player like Komarov for one or two years, he’s not likely to go for that without a compelling reason. Most players only take short deals if they think the amount on the next contract will go up, not down.
What the Leafs have to ask is do they want a player who brings what Komarov does — physicality, PK skills, elite fourth line defensive play, limited top-nine value — at a price that’s higher than Connor Brown, Zach Hyman or Matt Martin. How much better is Komarov than Martin, anyway?
My feeling is that Kasperi Kapanen has made Komarov redundant. The talk, and I heard some just the other day on TSN radio that Carl Grundstrom is the new Leo, seems years ahead of schedule, and a misread of Grundstrom’s game to some extent. But Kapanen has the PK skills (perhaps superior ones), the ability to play up or down the lineup, speed instead of hitting, tenacity in the corners, good forechecking, and a bit more scoring than Komarov can do in a good year, and a lot more than Martin offers.
Kapanen takes over most of the jobs Komarov does well, so unless the Leafs are pushed by roster realities into playing Kapanen as if he’s a top line winger, and I think that would be sub-optimal but not a disaster, then Komarov seems like a guy who should be given one hell of a going-away party.
The Leafs’ latest rental comes lower than Komarov on the list. He’s showing a projected AAV of $2,517,038 for three years. I’ll tell you, I’d make that deal any day. Tomas Plekanec on the Leafs could slot in as the 4C with William Nylander as the Bozak-esque sheltered 3C. Any time goals are really needed, Plekanec could move up to the third line, Nylander could move to Auston Matthews’ wing, and the bench could be shortened up by whatever fourth liners are left over.
That’s an ideal-world scheme to ease a player like Nylander into being a centre without giving up totally on what he brings as a winger. I love this plan so much, I have almost convinced myself it’s what the team should do at any cost.
But I think Plekanec is going to sign in Montréal where he would likely be the 3C all the time, and they might even give him more term than they should. I don’t think this dream can come true, but perhaps someone other than Plekanec can be found who will make a similar plan possible.
Roman Polak is well down the list, with an AAV of $1,963,758 for one year. He’s just ahead of the now-retired Francois Beauchemin. And that amount is 800,000 and change over what he got on a one-year deal coming off of injury last year.
Will someone pay him that? To play PK very well and third pairing minutes? I think there’d be offers. He shouldn’t play higher than that, and even when it seemed he was higher up the lineup this season, that was mostly illusion and people who like all-situations ice time stats when it makes the player they hate look like he’s playing more than he is.
Will the Leafs sign him again? I don’t know. I didn’t think they would this season. The Leafs have enough third-pairing defenders to hold a bulk discount sale, but never say never, I guess. Not many of the Leafs’ excess are right-shooters.
I think Dominic Moore was signed last year as an emergency last-minute, we can’t find anyone else, possible replacement for Eric Fehr. The fact that he supplanted Fehr, perhaps unfairly, was down to him making the most of his opportunity.
He didn’t add anything much beyond a few months of inflated shooting percentage. He was not horrible, but at his age, 37 right now, the model’s prediction of one year at $865,581 seems valid. I expect him to get offers again for the veteran’s courtesy amount of an even one million — now that you can bury that much in the AHL — and it won’t be the Leafs offering. Centres are hard to find, but the Leafs have to have grown beyond putting Moore in their lineup, or the big boss should be asking some hard questions of his staff.
William Nylander is the fifth person on Matt Cane’s list, which makes him likely to be paid more than almost all of the UFAs on the market. Ahead of him are John Tavares, Mark Stone, John Carlson and Evander Kane. The model produced AAVs at all contract lengths for Nylander that are about $1 million below Kane’s, until they converge on approximately $7 million at longer terms.
The model picked five years, but I’m not even considering that. The bridge deal for high-end first-time RFAs is a popular thing to talk about, almost in offer sheet territory for amount of verbiage relative to how often it’s really used, but a bridge deal only favours a team like Tampa who needed to shove Nikita Kucherov forcefully (or ask him nicely) under the salary cap for a few years.
I don’t see the fit for the Leafs on a shorter deal. Nylander might want to bet on himself commanding more in a few years, but that bet also includes the very real risk that injury could derail his plans.
One other consideration that shouldn’t be overlooked is that at Nylander’s level, his contract is a status symbol, and it puts him in a pecking order — not just relative to Matthews and Marner, but relative to David Pastrnak, Nicolaj Ehlers and any other similar player of a similar age.
He’ll get seven years, I think. Only Matthews is getting eight. And the AAV might not be quite $7 million, but I won’t bet heavily on it not being that either. $6.5 should be considered bargain basement.
Connor Carrick, who has limited NHL usage over the last two seasons is in a tough spot now that contract talks are due. The model has him at a two-year deal of $1,249,401. The one-year amount is just under $1 million.
He has arbitration rights, and if Kyle Dubas wants to be ruthless, he should elect arbitration on Carrick. Carrick has very little case to make beyond his shot differentials in sheltered usage. He’s not a PK man, he’s 24, he’s a coveted right-shot, but he is never going to get anything but emergency power play time on the Leafs. He’s considered a primarily offensive defenceman, yet there’s just one problem with that. He never gets points.
His career high at five-on-five was this season with nine. Roman Polak had 11. Carrick spent the bulk of his time on the ice playing behind the Bozak line or the Matthews line, so his on-ice goal differential is great, helped by a high on-ice save percentage. But he didn’t add much to the achieving of any of it beyond some seemingly real ability to drive play, although his offensive pace dropped a lot this season, which could simply be random chance.
It would be deeply ironic if the supposed analytics GM took a player to arbitration and argued against his Corsi or Expected Goals numbers, but I’d do it if Carrick won’t take a cheap deal. He seems to have topped out at alternate third pairing defender, and with Travis Dermott on the team, who needs his emergency power play skill? He’s not worth more than Josh Leivo money, and there’s no reason to overpay him even if it wouldn’t harm the team much to boost him up to something like the model predicts.
There are several RFAs who played none or limited NHL games on this list, and I’m skipping over them, because I don’t think the model has enough data to really calculate good answers on those players. The most interesting case of all them is Andreas Johnsson, and we’ll save that conversation for a post of its own.