Top 25 Under 25: Travis Dermott joins the party at #8

Travis Dermott has played 101 NHL games, split over the last two years. At 22 (turning 23 in December), he’s still at a point in his career where we can expect notable year to year improvement. Over the last two years, he’s anchored the Leafs third pair and driven it to very strong on-ice results. It’s fair to say his base level is an elite third pairing defensemen, and he has significant upside beyond that.

So why did he drop from #5 on last years ranking to #8 this year?

The Votes

The strong seasons of Andreas Johnsson and Kasperi Kapanen vaulted them ahead of him, and the arrival of Alexander Kerfoot means there’s another body to contend with in the group of “good but not elite bona fide NHLers”, below the star-level triumvirate of Auston Matthews, Mitch Marner, and William Nylander. Rasmus Sandin’s emergence hurts Dermott in these rankings as well, though with a three year age gap between the two left-shooting defensemen, they’re not quite in the same cohort.

On Dermott’s end, I think he also suffered from “more of the same” syndrome. We like to see players progress from season to season, particularly at a young age. Dermott ended the 2017/18 season anchoring a sheltered third pair with strong results. He spent the 2018/19 season doing largely the same thing. For fans who excited themselves over the possibility of Dermott stepping into a more meaningful role on a defence-starved team like the Leafs, it’s easy to see why last season may be considered a bit of a disappointment. His shoulder injury in the latter half of the season crippled him - even when he returned, he didn’t look 100%, and he eventually became part of a pairing with Jake Gardiner that is accurately described as two capable defencemen who were as healthy as the Antarctic ice sheet. To make things worse, his shoulder injury required offseason surgery that would reportedly keep him out of action for 6 months. This news broke in May, making it all but certain that Dermott will not be part of the Leafs opening day lineup.

Put all of this together, and Dermott’s slide is understandable. However, Dermott is still a tremendously valuable player who has the upside to be a significant contributor on a championship level team.

What type of player is he?

Dermott is a smooth-skating and agile defenseman who is comfortable with the puck on his stick. He plays an aggressive defensive style, particularly in the neutral zone, where he gets right up into the sweater of opposing forwards to disrupt them as they attempt to cross the Leafs’ blue line. Despite this, he’s been a positive penalty differential player throughout his career, which is a relative rarity for a defenseman (that said, this is a trait that is aided by the fact that he rarely faces high end forwards). He’s an active defenseman in the offensive zone, ranking 43rd among regular NHL defensemen last year in shot rate. To me, his calling card is his edgework, which is the core of his game both offensively and defensively. It’s what allows him to play so tight to opposing forwards, as well as create space for his own and his team’s offense.

What has he done?

Dermott is perhaps the quintessential example of a sheltered third-pairing defenceman who  puts up very strong on-ice results.

You can see the Leafs defense exists in tiers of playing time. Rielly is the constant, soaking up large amounts of 5v5 and 5v4 time throughout the year. Gardiner was keeping pace with Rielly at the start of the year, but his minutes were starting to get scaled back, and his injury obviously hurt his availability. Hainsey and Zaitsev were used in similar ways as the Leafs’ secondary defensemen, with a tilt towards defensive, 4v5, and lead protecting usage. Muzzin (helped by the injury to Gardiner) became the clear #2 to Rielly’s #1 after an acclimatization period.

Below them all was Dermott, who was the clear #5 for most of the year (and in the brief time with Muzzin included and all other defencemen healthy, the clear #6). There was a large gap above him - he was not really near the usage of anyone who played more than him. And there was an even more notable gap below him - none of the Leafs other third pairing options were close to his total TOI. Dermott started the year playing some 4v5 minutes, but this diminished over time, and the addition of Muzzin saw this experiment ended, more or less for good.

This pattern holds if we look at 5v5 time only.

Dermott spends the vast majority of his time on the third pair, as anyone watching the Leafs would agree with.

As you would expect, these third pairing minutes were ... rather cushy, in just about every way.

He faced incredibly easy competition, particularly with respect to forwards (note how low down he is on the y-axis, especially compared the Leafs’ top 4 defencemen)

He is used less in the defensive zone than any of the teammates playing more than him.

And finally, he is not as trusted in the important times of the game (particularly defensively) as the players above him in the lineup.

All of this is probably quite clear to fans who watched the Leafs last year, but it’s useful to contextualize his usage and note how clear the division is between the Leafs top four and bottom two. It’s not just a demarcation; they’re barely even playing the same game. In fairness to Dermott, his competition is fairly weak, but his teammates are too, and we would be remiss to not point that out. Dermott doesn’t get to spend his shifts stapled to Auston Matthews or John Tavares the way Morgan Rielly does.

Keeping this context in mind, we can look at the results the Leafs get when Dermott is on the ice. And... well, they’re pretty strong. Per NaturalStatTrick, Dermott had a 54.71% score and venue adjusted CF%, which is 26th in the league among defensemen with at least 500 5v5 minutes. Dermott isn’t quite as strong by xG% - only 51.14%, good for 79th among the same dataset, which is still nothing to sneeze at. Remember, he’s not playing with the Leafs stars very often, so a nice bit of offensive skill from Dermott still leaves the puck in the hands of Frederik Gauthier, rather than Auston Matthews. This is a contributor to his xG share lagging behind his shot share.

This xG note aside, Dermott is clearly succeeding in his current role. As mentioned earlier, his current level is that of a player who can do more than survive on a third pair — he can drive play and chances. But even a great third pair guy is just a third pair guy. Their value is inherently limited by their role. So we need to figure out not just what Dermott is now, but what he can be.

What is to come?

This is the hard part. Almost everything I said in the previous section could have been cut and pasted from what we were saying after the 2017/18 season, based on which fans wanted to see him deployed in a more significant role than we saw him actually used last season. Mike Babcock saw some criticism form certain corners of the Leafs fanbase for failing to find a more significant role for Dermott. In Babcock’s defense, the Leafs had two (in my eyes) superior left sided defencemen in Jake Gardiner and Morgan Rielly. On the other hand, Dermott wouldn’t be the only left shot in the world playing the right side. There is evidence that defensemen playing their off side are less effective (which is intuitive). However, Dermott would only have had to be better than Hainsey or Zaitsev to make this trade-off worth it. That’s a fairly low bar, in the eyes of many.

To defend Babcock on this, one can fairly point that Dermott is not the first defenseman playing low minutes against weak competition (and teammates) who is a Corsi superstar. In of itself, that doesn’t mean he will succeed when placed into the alternate world occupied by the Leafs top four. They’re very different roles with very different responsibilities. That is especially true when you consider that the players he would be replacing (Hainsey or Zaitsev) have notably more defensive usage than the rest of the Leafs top four, not only in terms of zone and competition, but leverage and game state as well.

On the other hand, the best indicator of someone being able to progress upwards in the lineup is success at lower levels. Yes, Dermott has had sheltered usage, but he’s also kicked ass in said usage. If we look at Isolated Threat and RAPM, two measures that attempt to isolate a player’s ability to drive expected goals, Dermott looks excellent.

Based on the last two years, Isolated Threat measures Dermott as 12% better than league average at driving expected goals for, and 1% better than league average at suppressing expected goals against. These are numbers that are in line with top pair defenders, not just top four.

Per EvolvingHockey’s RAPM, Dermott has been an elite defenseman at driving shot attempts, ranking 16th in the league over his career in Corsi RAPM (this method estimates that he created about 4.41 net shots per 60 minutes above an average defender). Again, he looks slightly worse if we look at xG RAPM, but he only falls down to 36th in the league - still comfortably in the realm of top pairing defenders (his estimated impact here is 0.16 expected goals more than an average defender per 60 minutes).

These methods are not perfect. It is fair to say that they are sometimes too kind to players exactly in Dermott’s situation (overqualified players beating up on easy usage). Nonetheless, I think there’s a valid argument that the Leafs should have tried to test Dermott out in a more significant role last year. Last October, it seemed that would be the case. Some predicted of Hainsey and Dermott would slowly swap places throughout the year, after the early parts of the season indicated a small uptick in playing time for Dermott. As we’ve already established, this didn’t happen. If we plot Dermott’s rolling 5v5 TOI, we see that he never consistently breaches 16.5 5v5 minutes per game.

As with many items, Babcock didn’t see things the way a lot of Leafs Twitter did, and in any case, Dermott’s shoulder injury prevented him from seeing his playing time increase after Gardiner went down. This year, his recovery from offseason surgery will see him miss training camp, and as a result, critical low-pressure experimentation time. When he does return, many will be hoping that he is played above Cody Ceci as the 2RD. I am skeptical that will occur. I think it’s fairly likely we see year three of the same story with Dermott, where he’s stuck behind two unequivocally more established players on the left side (Rielly and Muzzin), and behind a right-sided defenceman that is a clear weak spot on an otherwise strong team (Ceci).

The question of where Dermott plays this year occurs against a backdrop of uncertainty over his next contract. Dermott is exactly the type of player who gets undervalued in free agency. He’s stuck behind better options at 5v4, so he doesn’t put up a huge amount of points. He’s not played heavily, so TOI isn’t a point in his favour. His best measurable attribute is how he’s driven play. In that way, there could be some significant upside in trying to extend him now. He hasn’t proven that he is a bona fide top four defenceman, and as such, one would think an extension would reflect that. On a team like the Leafs, where their elite players are being paid what they are worth (or more, in some cases), being able to find contract value for a player that can meaningfully contribute in the future is important. The risk the Leafs take in that scenario is whether Dermott actually develops into that player. If he does, extending him would be a source of immense value for Toronto going forward. If he doesn’t, the Leafs would have committed notable resources to a player who ultimately doesn’t play a major role on a contender.

He’s done almost everything he could have with his play to show that he’s ready for a bigger role. The question is whether he gets it, and what happens when he does. Your guess is as good as mine on that front.

What other panelists thought

Brigstew: If there’s ever a year for Dermott to win a spot in the top four, it’s going to be this year. On the one hand Muzzin and Rielly are still blocking him on the left side. On the other hand he has experience playing on the right and while I don’t think Ceci will be as bad as we all feared, I also would like to think that Dermott should be good enough for Ceci to not be a difficult hurdle to clear. Not if he wants to ever really be a top 4 guy one day. He has the nice looking fancy stats in a more limited third pairing role, so this year I want to see him take this opportunity and run with it.

Katya: Travis Dermott has been a roller coaster ride for me. I was totally mystified to what people saw in him in the AHL. I saw a player who did boneheaded things under pressure, and who had more desire than skill. In the NHL, he was immediately, and obviously good enough for the job he’d been given, and I was thrilled, but now, after some sober second thought about his play, I’m left wondering if he can climb much higher.

Usage is key when evaluating defenders on the Leafs. They have very clearly defined roles, and they are deployed by strict rules. The Leafs change defenders when the play is heading up the ice, with the Leafs in possesion. The third pair (Dermott and Ozhiganov) were the most extreme examples of this. This isn’t about zone starts, this is about deployment in an on-the-fly change (which is most of them) that means the defender is almost never tasked with defending right away. They have to first fail at sustaining possesion. Oz and Dermott managed a 50% on-ice Expected Goal differential in that usage. This does not scream out: give them a harder job.

There’s more to this story than just when Dermott hit the ice. The Leafs change their forwards on the backcheck, so it’s extremely easy to snatch the poorest defensive forwards off the ice when the weakest defending defenders are heading back the wrong way.  For every minute Dermott spent with a poorer offensive forward, he got some portion of one well away from the Auston Matthews line defensively. Before his late-season injury, Dermott was getting more minutes than Oz -- notably some shifts with Jake Gardiner -- but he wasn’t really getting harder minutes.

Can he handle harder minutes? The truth is, we still don’t know, but if you trust his raw Corsi stats, you will have fallen for an illusion created by deployment, and not reality. The most likely answer is he won’t make you cry on the second pair, but he won’t make you say, “Finally, a real defenceman!” like Jake Muzzin does either.