We’re just about at the quarter point of the season, 19 games in, and the Leafs have a four-day lull between games. It’s a good time to take stock and see how our players have fared. Fulemin looked at the forwards on Monday, and I examined the defense on Tuesday. Today, we’ll talk goalies and coaching.
The grading scale is as follows:
A: Exceeded expectations
B: Met expectations
C: Slightly short of expectations
D: Serious disappointment
F: Everything is dark and sad
Stats are 5v5 and from Corsica unless otherwise noted.
Frederik Andersen: C-
Role: Undisputed #1 goalie
Andersen’s play to start the season has sparked a large discussion about a goaltender’s culpability in a side leaking goals, relative to that of the skaters in front of him surrendering the chances. His raw stats are not pretty. His 5v5 save percentage of 90% puts him 38th out of 46 goalies with more than 200 5v5 minutes. The only players behind him are Vegas’ 4th string goalie, Arizona’s duo of sadness (one of whom has since been traded), Winnipeg’s backup, Washington’s backup, and (hilariously) Tuukka Rask and Carey Price. Simply put, he hasn’t done a great job of keeping pucks out of the net, relative to his peers.
Does this mean that Andersen is now bad? No, definitely not. He has a solid track record of being a league average goalie by save percentage, and there’s no indication that this is anything beyond a short-term blip. At the same time, we are grading on what players have done, and he hasn’t done all that much on the positive side of the ledger.
However, it can very reasonably be argued that the only reason his numbers look so bad is because the Leafs routinely hang him out to dry.
The scales can make this seem more extreme than it is, but there’s no denying the Leafs allow a lot of shots directly from the high slot, and a lot of dangerous shots in general. This is backed up by the numbers too. Corsica has the Leafs as 16th best in the league in FA/60 (unblocked shots against). If you adjust for shot quality, however, they rank 29th in the league. Essentially, the Leafs give up an average amount of unblocked shots, but the shots that they do allow are of very high danger, which results in them being a poor defensive squad by expected goals.
What this has resulted in is a lot of goals against where it’s hard to blame Freddie a whole lot for letting them in. They’re high quality chances, the kind that get scored relatively often in hockey, and in many cases, there’s an obvious error by one (or more) of the five skaters in front of Andersen preceding it. Consequently, there’s been a divide in people assigning blame to Andersen for his poor numbers. On one hand, it’s unreasonable to expect him to turn in a save percentage that will lead the league when the Leafs have comedic defensive breakdowns on a game-by-game basis. On the other, it is equally unreasonable to absolve him of all blame once those breakdowns occur. Not every defensive breakdown turns into a goal. Even if a shot has a 90% chance of going in (which is incredibly rare in hockey - a ‘good’ chance might only have a 20-30% chance of being a goal), you need to save it 10% of the time, or you’re going to be below average.
So with that in mind, lets try and find a balance between the two. We (or rather, Corsica) can compute the expected save percentage of a set of shots. By comparing Andersen’s numbers to what we would expect an average goalie to save with the same workload, we can get a better sense of how to appraise his play.
Unsurprisingly, this process makes Andersen look far better over the first quarter of the season than his raw numbers might indicate. Still not good, to be sure. But better. His dSv% (difference between actual save percentage and expected save percentage) is -1.1%. Essentially, this means that an ‘average’ goalie would have saved 91.1% of the total shots faced, given Andersen’s workload, while Freddie only saved 90%. This ranks 32nd of the 46 goaltenders who meet the TOI cutoff established earlier.
I think it’s important to note here: I am not inferring any predictive ability from the models used to generate the expected goals (and consequently, expected save percentage). This is purely descriptive. I’m focusing on what Andersen has done, versus what Andersen is.
So all of this leaves us here:
- Andersen’s numbers have not been good.
- The Leafs put him in a spot that is very difficult to succeed in
- Accounting for that, he appears to have done relatively poorly on aggregate, though not as bad as it seems at first glance. It must be said that your confidence in this statement will be wholly dependent on how you feel about the shot quality model we’re using to adjust his workload. Corsica’s is not predictive, but it has been shown to be an accurate descriptor, so I am satisfied with it in this case.
Now, does this mean that Andersen is bad, and that the Leafs are doomed for hitching their wagon to him? No, not at all. He has a track record of being a solid goalie, and last year, some metrics have him as one of the best in the league (once adjusting for shot quality). After a brutal start, he’s already had some solid games recently, and throughout the season, he has been above average on the PK (which helps his grade a little). Going forward, there’s no reason to think he isn’t still the same guy we thought he was.
Curtis McElhinney: A
Role: Chief Clipboard Operator, Real good pro, Nonthreatening backup
Not a lot to say here. McElhinney has had one ugly start and two good ones. Do I think he’s a particularly good goalie? No. Do I think he’s the second best goalie in the Leafs organization? No. Do I feel at all confident when watching him? No.
But he’s gotten very solid results so far, so he gets an A.
Mike Babcock (and his staff): B+
I complain a LOT about some of the moves Mike Babcock and his staff make, but rest assured, they get a lot of things right, some of which we take for granted. Special teams are an obvious example.
Teams more blessed with offensive talent than the Leafs have had bad power plays. Not something we have to worry about. The coaching staff has created a system that generates quality shots when up a man, and a lot of them to boot.
The Leafs are 9th in the league in CF/60 when up 5v4. They’re 4th in GF/60, and are 2nd in xGF/60. As I said, they generate a lot of shots, and a lot of quality shots. It’s a great combination. Impressively, their top unit doesn’t have their top offensive player on it. While the unit of Kadri/Bozak/van Riemsdyk/Marner/Rielly is by no means underpowered, it is a testament to the system that they run, and their own talent, that they are one of the best units in the league. This was discussed nicely in some detail by Tyler Dellow of The Athletic.
Now, let’s turn to the PK. Toronto ranks 15th in CA/60, 13th in GA/60, and 11th in xGA/60. Not an elite group, but a reasonably decent one, especially with how bad Toronto started off the year in this regard. Worth noting is that Nikita Zaitsev and Ron Hainsey have played the lion’s share of these minutes - they deserve a lot of credit for that, though it is uncertain to me how sustainable it is to lean so heavily on just two defensemen for PK duties.
There are two things that stop the coaching staff from getting an A. The first is the slightly suboptimal lineup choices that are made from time to time. I actually don’t ding them much for this - I tend to believe that coaches know what they’re doing. However, in some cases, it seems obvious to me that there is no justification for a particular move. Playing Roman Polak is one of them. Playing Matt Martin arguably is as well, but he is at least a capable fourth liner who plays fourth line minutes.
The other is potentially more worrying. Over the last few weeks, the Leafs have seen their possession stats slip. Once league leaders in CF%, they are now bang average, relative to the rest of the league. Teams have been able to stymie Toronto’s (theoretically) high flying offense, at least in terms of generating shot quantity. What results is a lot of the Leafs playing in their own zone. As we covered when discussing Andersen, they’re not good at that.
So far, there hasn’t seemed to be a real solution to this issue. To be sure, this is 100% a problem that is split between players and coaches. It is also mitigated by the fact that the Leafs tend to shoot from the dangerous areas of the ice.
However, the Leafs need to get into the offensive zone in order to wreak havoc there. Over the last few games, they haven’t been quite as good in that regard, which means I can’t give the coaching staff an A.
In a sense, this is a result of having good coaches. My expectations for Babcock are higher than they would be for just about every coach in the league, and he is doing a lot more right than he is wrong. As much as fans may complain about the decisions he makes on the margins, he gets the big stuff right, and after years of coaching ineptitude, that is priceless.