At the 2015 NHL Draft, the Maple Leafs and Oilers made what appeared to be a non-descript trade: Toronto sent a fourth-rounder and prospect Brad Ross to Edmonton for 23-year-old defenceman Martin Marincin.
Fast-forward to today and Marincin is skating on Toronto’s top pair, riding shotgun with budding star Morgan Rielly. Despite some very intriguing results, some have already dubbed this a “failed experiment.” The perception that Marincin has made a number of bad giveaways that directly led to losses exemplifies for many what is wrong with his game.
This is not new territory for Marincin observers. The Oilers traded him for largely this reason: a perception that he wasn’t contributing in a fashion commensurate with his size. While Marincin has the size many NHL scouts drool over, he’s seen as soft in his own end and not able to contribute enough on offence to offset that shortcoming.
In essence what we are describing is another case where the “eye test” falls short in recognizing a measure of effectiveness. Interestingly the Leafs, a team that is becoming increasingly analytically inclined, seem to see a lot of promise in Marincin.
This week, coach Mike Babcock said the team thinks Marincin “has a lot of potential” when he was talking about promoting him to the top pair with Rielly.
Every statistical measure we can examine tells us Marincin is an extremely effective defensive defender. He is the Leafs’ most effective shot attempt suppressor when it comes to possession. He is on the ice for the fewest goals against per 60 minutes at 5v5 of any Leafs defender. He is on the ice for the fewest scoring chances per 60 minutes at 5v5 of any Leafs defender. He’s on for fewer of those “highlight reel” plays against than his teammates, too.
The NHL website logs individual scoring chance plays for users to queue up and watch on video during or after the game. These plays are marked in the game files and thus it is possible to track how frequently a player is on the ice for a highlight reel play for or against, giving us an interesting measure of what the eye test should actually show. If our memories really were accurate recording devices that tracked every positive and negative event you’d think it would align with these tracked plays.
Emmanuel Perry, the developer behind the new hockey statistical website Corsica.Hockey has logged all of the Highlight Reel (HL) events over the past three years using the NHL’s game files and it paints an interesting picture of who the eye test should really be celebrating on offence (HLF) and who it should be blaming on defence (HLA).
Two things become clear when we look at these events as rates: 1) Marincin should not be considered an offensive dynamo, which makes sense, and 2) His perceived failures via the eye test on defence are grossly disproportionate to his actual negative involvement.
| Highlight Reel Plays | |||||||
|---|---|---|---|---|---|---|---|
| Player | iHL | HLF | HLA | iHL60 | HLF60 | HLA60 | HLA60 Percentile |
| Marincin | 16 | 119 | 141 | 1.38 | 10.23 | 12.12 | 7.40% |
| Gardiner | 49 | 308 | 261 | 2.38 | 14.96 | 12.68 | 13.00% |
| Phaneuf | 80 | 364 | 318 | 3.3 | 15.01 | 13.11 | 17.20% |
| Polak | 32 | 211 | 287 | 1.57 | 10.35 | 14.08 | 34.20% |
| Rielly | 61 | 293 | 382 | 2.46 | 11.83 | 15.42 | 59.40% |
| Hunwick | 38 | 243 | 401 | 1.68 | 10.76 | 17.76 | 88.20% |
The Percentile column shows what percentage of NHL defenders with 200+ minutes of TOI have allowed a lower rate of Highlight Reel Events Against per 60 minutes of play over the past three seasons.
Marincin is only on the ice for 12.12 Highlight Reel Events Against his team for every 60 minutes he is on the ice. Similar defenders by this measure would be Brian Campbell (11.73 HLA60), Hampus Lindholm (11.86 HLA60), Colton Parayko (11.89 HLA60), Sami Vatanen (12.14 HLA60), Seth Jones (12.15 HLA60), and Mark Giordano (12.20 HLA60). Not exactly skaters who anyone would describe as poor on defence.
There is actually very little concrete evidence to support the idea Marincin gives up a wildly disproportionate number of top end scoring chances to the opposition. If we look at 5v5 shot attempt and goal-based metrics for this season a very similar picture is painted of Marincin being an elite defensive defender who makes a positive impact by suppressing shot attempts against.
| Defensive Metrics | |||||||
|---|---|---|---|---|---|---|---|
| Player | TOI | GA60 | xGA60 | HDSCA60 | SCA60 | Shot Att A60 | Shot Atts Suppressed |
| Marincin | 618.9 | 2.04 | 1.97 | 9.69 | 24.14 | 48.57 | -49.51 |
| Polak | 881.18 | 1.97 | 2.16 | 9.87 | 26.9 | 54.2 | -53.12 |
| Gardiner | 1055.82 | 2.33 | 2.37 | 10.63 | 26.88 | 52.45 | -37.31 |
| Phaneuf | 853.25 | 2.11 | 2.32 | 10.34 | 27.28 | 56.75 | 22.57 |
| Rielly | 1166.43 | 2.88 | 2.59 | 12.14 | 29.89 | 58.95 | 74.9 |
| Hunwick | 1058.41 | 2.66 | 2.71 | 12.87 | 30.22 | 61.62 | 111.16 |
An xGA60 rating under 2.00 puts Marincin in the top 13 defenders in the entire NHL. His peers in this regard would include the likes of Marc-Edouard Vlasic, Anton Stralman, Justin Braun, Mattias Ekholm, Seth Jones and Hampus Lindholm.
THE LESSON
Human memory is amazingly amorphous and flexible – often to negative outcomes when we think about player evaluation. We fixate on relatively rare events and blow them up to much greater significance than is warranted, usually in an effort to confirm our preconceptions about the player we are watching.
Over the course of this season Marincin has exceeded defensive expectations as indicated by every measure we have at our disposal. He is only 24, and if his development can continue he stands a decent chance of developing into an excellent shut down defender. The final quarter of this season will give ample opportunity to see how Marincin handles a higher minute load in a top pairing capacity, but so far this year the data would suggest he is capable of meeting the challenge.
