Defensive Tandem Performance

21-Ference-6

It’s a sure thing that Peter Chiarelli plans to overhaul the Oilers’ defence, and we’ve seen him at work already. There are, however, five holdovers from last year’s group, and we can probably learn something from what kinds of players the incumbents failed/succeeded with last season.

How it Works

  Schultz Petry Fayne Nikitin Ference
Klefbom 726.8 51.4 104.5 52.8 218.2 43.3
Ference 154.9 42.4 694.5 46.0 83.1 44.5 NA NA
Nikitin 175.6 42.7 75.3 50.7 294.6 49.6 NA NA
Marincin 75.5 56.9 331.5 46.3 62.3 54.9
Aulie 82.8 50.7 103.7 42.9 68.6 39.0
  TOI CF% TOI CF% TOI CF% TOI CF% TOI CF%

Basically, the idea here is to identify how well defensive tandems worked together vs. how well they should have been expected to work together.

To that end, I’ve used the numbers posted by centres as proxies for their lines, and gone through the tables at stats.hockeyanalysis.com to see how much time each defenceman spent with each centre. So, as an example, a defenceman who spent half his time with Boyd Gordon (43.5 percent Corsi) and Ryan Nugent-Hopkins (49.9 percent Corsi) would be expected to produce a number halfway between the two (46.7 percent Corsi).

There are some limitations to this method. We’re pretty much stuck with Corsi as our measuring stick because goals are too infrequent for samples this small. We aren’t totally allowing for quality of competition or zone starts, though some of that obviously overlaps with playing with a certain group of forwards (e.g. defencemen who spend a lot of time with Gordon generally have severe zone starts, but that’s allowed for somewhat in Gordon’s wretched Corsi).

But enough of that; let’s get to the result.

Findings

Left D Right D Expected Actual Diff.
Marincin Schultz 48.3 56.9 8.6
Marincin Nikitin 48.2 54.9 6.7
Klefbom Petry 48.5 52.8 4.3
Klefbom Schultz 48.6 51.4 2.8
Nikitin Petry 48.6 50.7 2.1
Aulie Schultz 48.6 50.7 2.1
Nikitin Fayne 48.3 49.6 1.3
Marincin Fayne 48.0 46.3 -1.7
Ference  Petry 48.5 46.0 -2.5
Ference  Fayne 48.2 44.5 -3.7
Klefbom Fayne 48.3 43.3 -5.0
Aulie Fayne 48.3 42.9 -5.4
Nikitin Schultz 48.6 42.7 -5.9
Ference  Schultz 48.6 42.4 -6.2
Aulie Ference 48.4 39.0 -9.4

The chart above shows the performance (by Corsi) actually achieved by each defensive tandem vs. their expected performance based on the forwards they played with. It’s an interesting tool; what does it say about the individual defenders involved?

22-Aulie-1

Keith Aulie probably isn’t an NHL player, but the Oilers hardly put him in a position to succeed, either. Given his obvious difficulty in handling the puck, it seems abundantly clear that he needed a partner who could be a primary puck-mover. When he had one (Schultz, whose weaknesses are somewhat compensated for by Aulie) the results were decent, admittedly in a small amount of time together. With Fayne and Ference, though, he was a boat anchor.

5-Fayne-5

Mark Fayne doesn’t impress too much here, and this exercise has me slightly revising my opinion of him downward. He’s obviously a complementary player, and one who needs a puckmover on his pairing; his results with Aulie and Ference show that pretty clearly. His work with Klefbom was really disappointing, though, since Klefbom would seem to be exactly the kind of player that Fayne would do well with and it just didn’t happen. Fayne was better with Nikitin and Marincin, two players capable of moving the puck.

21-Ference-3

Andrew Ference looks bad here, though some of it is likely related to usage. He and Petry had a pretty tough defensive zone starts assignment and he and Fayne had an absolutely brutal one (48.5 percent of all shifts starting with a faceoff, including neutral zone shifts, took place in the defensive zone). There’s no excuse for his disastrous performance with Schultz, though, and his time on the right side with Aulie as his partner was a complete trainwreck. Nothing here screams “actual NHL defenceman”.

84-Klefbom-5

Oscar Klefbom’s work with Schultz outshines that done by any of the Oilers’ veteran left side defencemen. Schultz was a disaster with Ference and Nikitin, but with Klefbom things worked out nicely. They did have the advantage of a pretty soft zone start (the tandem was twice as likely to start in the offensive zone as in the defensive zone, and 40 percent of all shifts starting with a faceoff took place in the opposition end). He also did well with Petry, though oddly he and Fayne struggled together.

85-Marincin-4

Martin Marincin’s handling remains baffling in hindsight. All of his partners fared better with him. Schultz and Nikitin in particular struggled in certain situations but looked worlds better paired with the lanky Slovak; even Fayne in a tough minutes role improved next to Marincin. Instead, the Oilers constantly exhibited a preference for inferior veterans, and shipped away the player for a marginal return at this summer’s draft. There probably wasn’t room for him long-term on the left side, but if the team had been a little more willing to use him he could have made these last two seasons a little better.

86-Nikitin-3

Nikita Nikitin comes across better here than I really expected. He and Marincin played pretty well together, which is a nice reminder that Nikitin’s most successful time in the NHL came on the right side with a skating, puck-moving partner. He and Petry were also surprisingly good together, though the caveat in both cases is limited minutes.

The limited minutes caveat doesn’t apply to his partnerships with Fayne and Schultz. Nikitin and Schultz were a total disaster and probably shouldn’t have lasted together as long as they did. Nikitin and Fayne were shockingly competent by this metric, though, and might end up seeing time together again.

2-Petry-4

Jeff Petry is going to be missed. Ference posted his best numbers all season playing alongside Petry. Nikitin actually played better than the team average alongside Petry. Klefbom gets a lot of attention for his work with Schultz, but his work with Petry was even better. He was the rising tide that lifted all partners, and he did this for years on the Oilers blue line, a fact the team’s former general managers were sadly oblivious to.

19-Schultz-11

Justin Schultz is one of the players who surprised me in this analysis, and I come away thinking a little better of him. We know he had success playing with Klefbom, and that’s even after accounting for linemates, but he did a surprisingly decent job next to both Aulie and Marincin. His numbers with Marincin are always good, year-after-year, in small samples and year-after-year the coaches refused to let those samples get larger. This was an awfully good tandem in the AHL in 2012-13 when both players were rookies and it always astounded me that it never really got tried consistently in the majors.

In contrast, Schultz was a total disaster with both Nikitin and Ference. It might be nothing, but it’s interesting to me how well Schultz played when paired with a defenceman who can skate (even Aulie was pretty mobile) and how wretched he was with slower guys.

Lessons for 2015-16

19-Schultz-1

What I take away from this:

  • Schultz looks like he needs a skating partner (his success with the reasonably mobile Aulie suggests this); in other words keep him far, far away from Ference and Nikitin.
  • Klefbom isn’t terribly needy, but probably would work best with a puck-mover; so avoid Gryba and Fayne.
  • Fayne’s past history in New Jersey suggests he needs a puck-mover to enjoy success; unfortunately his recent history indicates that Klefbom isn’t the guy.
  • Nikitin’s going to be an interesting guy to watch this year, and I wonder if we’ll see a bounce-back campaign. He can slot in on either side and away from Schultz he really wasn’t bad.
  • Ference probably shouldn’t be in the starting six.

If we assume an all-veteran seven-man defensive unit, I’d be tempted to start out Sekera-Fayne (owing to Fayne’s success with Greene in past years), Klefbom-Schultz (if it ain’t broke…) and Nikitin-Gryba with Ference in the No. 7 slot.

If Griffin Reinhart makes the team, I’d suggest the same pairings to start with Reinhart on the third pairing and Nikitin/Gryba rotating as his right-side partner. On the other hand, if Darnell Nurse is ready to go, it’s tempting to run Klefbom-Sekera on the top pair and then use Nurse-Schultz as an offensive pairing and Nikitin-Fayne as a defensive pairing.

One thing I really wonder if we’ll see is Schultz paired with Sekera, even just on a trial basis. Schultz has never played with a veteran like Sekera, who in a lot of ways is what a perfect-world version of a developed Marincin would look like. That pairing could play a ton of minutes and Schultz would certainly never have a better shot at success.

  • Seriously, do you want a season with any of Ference, Schultz and Nikitin playing?

    Chiarelli missed most of his chances to flush these players.

    But he has enough forwards available to still trade his way to a competent defense. In fact, he has quality forwards sufficient to acquire a real #1.
    But the will has to be there.

    No point in not starting Nurse and Reinhart from day 1 of this season.

  • bradleypi

    It looks to me like the CF% tells us very little about the quality or the pairings of the d-men. Zone starts will definitely tell you why most of those numbers are the way they are. Sorry Jonathon but I think this article accomplishes very little.

  • Ndustry

    Ahh, it’s just Corsi! If you look at the Edm Journals Nielson and mistakes and contributions to scoring chances things look different. Ference looks better when you count points again, Marincin is a statistical anomaly who went under appreciated, but to defend getting rid of him despite his impact on Corsi he had a negative impact on offense and brought down the offensive quality of play and on ice shooting % of even our most potent offensive players, Hall and Eberle.

    Fayne is highly touted but in this view seems to back up his ice time he’s been 6th or less D in ice time his whole career. Nikitin is below projections and expected performance so it is likely he will balance out to a more reasonable trajectory. Schultz has the the 8th highest GF%Rel TM for defense playing over 500 mins in the league, so despite his current point totals he’s showing a high potential to produce even more and he’s involved in some very high quality on ice offensive play.

    Klefbom has impressed but he’s likely to be overrun in the Depth chart by both Nurse and Reinhart, unless he projects high offensively to which he has to compete with Sekera and Schltz plus even Niktin had a productive offensive year before too.

    Once again it’s just Corsi though, GF60 GA60 CF60 CA60 the Edm Journals Neilson ratings, Ryan Stimsons passing project for Contributed Corsi and Zone entry and most importantly deployment aren’t to be overlooked, heck even the old fashioned +/- has some value too!