Photo Credit: John E. Sokolowski / USA Today Sports

Hunwick, Polak, and their quantifiable improvement

There’s been constant discussion over the course of this season on how to improve this Maple Leafs team on the margins. The areas of the game where a competitive advantage can be gained are in areas where the rest of the league is complacent. For a while, these areas for the Leafs were the backup goaltending, the 4th line, and the bottom pairing. With the acquisitions of Brian Boyle and Curtis McElhinney, two of the aforementioned three seem to have been solved. A third acquisition, Alexey Marchenko, was inferred as being part of a solution to the remaining issue. But with Connor Carrick’s injury, Marchenko has moved into that pairing with Gardiner. And thus, the bottom pairing of Hunwick and Polak remain.

There can be something said of the idea of moving on from unrestricted free agents at the deadline to get future value in the way of draft picks in return. As both Hunwick and Polak are such UFAs at the end of this season, the Leafs either chose not do so to or were unable to. So we’re stuck with them for now. But maybe – just maybe – that isn’t so bad.


Before we analyze the predictive metrics and recent results, it’s important to provide context in the way of Quality of Teammates (QoT) and Quality of Competition (QoC). This is provided in a hopefully easy-to-digest usage chart below.

Another D-core (the Chicago Blackhawks) is brought in in order to stabilize the axes a bit, and provide some visual reference. The chart above is from Corsica.hockey. This isn’t a typical usage chart, so please note the axes do not include zone starts. This was done because we’re going to be using zone-start adjusted numbers in the data analysis section below, so the actual numbers for zone starts aren’t necessary. The colour bubble in this case is GF% for results thus far.

On the chart we have 6 distinct groups:

  • High QoT, High QoC – Nikita Zaitsev, Niklas Hjalmarsson, Duncan Keith
  • High QoT, Medium QoC – Brent Seabrook, Brian Campbell
  • Medium QoT, High QoC – Morgan Rielly
  • Medium QoT, Medium QoT – Martin Marincin, Connor Carrick, Matt Hunwick, Roman Polak, Trevor Van Riemdsyk
  • Medium QoT, Low QoC – Michal Rozsival
  • Low QoT, Medium QoC – Jake Gardiner, Alexey Marchenko, Michal Kempny

There’s nothing especially interesting to gain from Matt Hunwick and Roman Polak being in the medium-medium cluster.  However, Gardiner and Marchenko being in the low-medium group should make us give them a small benefit of the doubt when analyzing their numbers. Additionally, Rielly and Zaitsev facing the high QoC should be a reasonable, but not all-forgiving, excuse for them.


Since the change of the calendar, Hunwick and Polak both have been much improved in shot attempt numbers. You can see the trend for their CF% below:

The graph above shows a steady uptick for the pairing not-so-affectionately referred to as “Hunlak”. It wouldn’t be unreasonable to think these are just team effects, so here are the actual numbers:

Player TOI CF% Rel.CF% CF% RelTM xGF% Rel.xGF%
MATT HUNWICK 437.92 51.88 1.91 -2.5 52.36 3.31
JAKE GARDINER 554.22 51.16 0.68 3.8 51.43 2.29
ROMAN POLAK 458.55 51.01 0.65 -3.6 51.56 2.53
NIKITA ZAITSEV 546.44 50.5 -0.39 -1.3 49.71 -0.48
CONNOR CARRICK 356.9 50.9 -0.53 3.5 47.39 -3.84
ALEXEY MARCHENKO 235.71 47.05 -2.48 -2.0 50.24 3.14
MORGAN RIELLY 430.66 48.85 -2.74 -0.6 46.36 -4.83

All data is 5v5. CF%RelTM is for the whole season, unadjusted, and retrieved from stats.hockeyanalysis.com. The remaining statistics are since January 1st, 2017, Score-Zone-and-Venue adjusted, and were retrieved from Corsica.hockey.

Look at Matt Hunwick. Since January 1 he leads the team in both rel CF% and rel xGF%, and Roman Polak isn’t far behind. However, for the season, Hunwick and Polak are the worst of the group in the RelTM numbers for the season. One possible explanation is that, since Jan 1, Polak and Hunwick have had much higher QoT in the period since Jan 1. Unfortunately, that data isn’t available in Corsica’s Custom Query section for confirmation of that hypothesis. But even if that were true, the difference in playing with better Leafs forwards than previously isn’t enough to explain their better numbers. They’ve just been better. Good, even. Which is significantly encouraging for their probable continued usage throughout the remained of the Leafs’ playoff push.

How Babcock had the stomach to keep this pairing together as long as he did is baffling. But at this point it’s looking like the decision is really paying off. It wouldn’t be egregious to suggest Hunlak have been the Leafs’ best-performing pairing this calendar year.


With the addition of Marchenko, Connor Carrick likely returning from injury sometime soon, and Martin Marincin putting in a good showing recently, a tough decision will have to be made. Whether to keep Hunlak together, or let some combination of Marincin, Marchenko and Carrick play, will be an interesting discussion. From the data above, it seems the ideal depth chart (who is most deserving of playing, not what pairing they should be on) would be:

Left side:                                                                                          Right side:               

Gardiner                                                                                              Zaitsev

Rielly                                                                                                    Polak

Hunwick                                                                                              Carrick

Marincin                                                                                              Rielly

Carrick                                                                                                 Marincin


The big detractor for Hunwick and Polak (and attractor for Gardiner and Carrick) is the RelTM numbers, which suggest that Hunwick and Polak have been significantly worse than the players they’ve been on the ice with this season. However, with their positive effect on the team (Rel.CF% and Rel.xGF% since Jan 1, it would be totally unfair to take them out of the lineup at this point.

Despite the proposed depth chart above, the best thing to do at this point in the season is to leave the pairings as they are. The biggest change that could help this defense core perform better is to stop utilizing Rielly-Zaitsev as a shutdown pairing, and instead spread those minutes across the three pairings. None of the three pairings are good enough defensively to be a strict shutdown pair. So instead of bogging down one pair, it might be best to let the team shut the top lines down by committee.

Either way, the conclusion here is that it’s time to stop bemoaning the use of Hunwick and Polak. In the words of Leafshub’s Jude Macdonald, “Don’t @ me about Hunlak.” They’re proving that right now, they’re good enough to play. Until that changes, it’s in Mike Babcock’s best interest to avoid fixing something that isn’t broken.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Stan Smith

    While “Hunlak” posted terrible Corsi for most of the season they have led the team in GA per 60, 5 on 5 the whole season. As I have stated numerous times, while the Leafs have issues on defence, Hunwick and Polak as the third pair is not one of them. It’s more like they have one third pair, and two decent second pairings, but lack a good first pair.

    • killerkash

      You’re on the right track Stan but I think we have 2 third line pairings and a second in Carrick and Gardner. Rielly and Zaitsev have looked god awful the last couple of games which makes it so very obvious the need for a couple of good, experienced defencemen in the off season. A guy like Karl Alzner to play with Rielly as a top pairing and another guy like Johnny Oduya to play on a third pairing with Zaitsev. That would solidify the defence to a point where only a better defenceman could break into the lineup on a steady basis and would also give the Leafs some depth back there with one of Marincin or Marchenko as the 7th defenceman and the other who may be taken in the expansion draft.

  • LeafsRamblings

    Jude Macdonald also said that Polak was better than Shattenkirk, so while the rest of this piece is quite sound and reasonable, this is essentially the equivalent of nailing a job interview and then pooping on the floor before you leave.