Splitting Up the Second Line: A Rebuttal

This morning, Sunil Agnihotri (@sunilagni) wrote a quick post at CopperNBlue arguing some caution in splitting up the second line.

As always, Sunil makes some excellent points and I suggest you click the link to read his points. I’ll only provide a quick summary to say that the line of Nugent-Hopkins, Lucic and Eberle have done well via shot attempts and expected goals so far in these playoffs while taking on the toughest competition. This would imply that while the results haven’t been there, the process is sound. If the process is sound, the goals are more likely to come.

Generally, I agree with this type of analysis. Goals are a rare event and while they are ultimately the most important result, sometimes you can make a more informed decision by analyzing process than by analyzing result. To that end, I wrote a piece singing the praises of Ryan Nugent-Hopkins after the San Jose Sharks series using the same basic approach.

However, I wanted to point out a couple things in response to Sunil’s post that ended up being too long for either a comment or a Twitter tweet-storm. Hence this post.

Analysis In The Playoffs

The playoffs are a tricky beast to analyze. It is, by definition, a collection of small sample size performances. Hockey is exciting precisely because anything can happen in a game. Some times Jake Allen is lights out and your team wins even though Minnesota has outplayed you for the series, for example. To that end, Lowetide also made a valid point in his piece this morning, writing about how “a cold streak for an offensive player can be devastating to player and fanbase” and urged caution in ripping Eberle apart.

However, beyond just small sample sizes, the playoffs are different because of the nature of the series. Lumping together 1st round and 2nd round performance increases sample size, but it also may be invalid given the competition is entirely different. That is unlike the regular season where more games generally allows the quality of competition to even out. You play some games against the Colorados of the world and some games against the Chicagos of the world.

In the regular season there is, of course, advance scouting and tactical matchups. However, my guess is that coaches are less likely to drastically change their deployment from game to game. It would be confusing for the players. You make some adjustments to the team you face, but you know that in 24 hours you’re going to facing a brand new team and might not have a full practice in between them if you want to modify tactics further.

That’s not true in the playoffs. In the playoffs you’ve got practices between every game. You’ve got time to focus your gameplan on ONE opponent for a series of games and optimize your gameplan against that ONE opponent.

Get To The Point Wheat!

My point in all this is that Sunil’s argument is based around the idea that you can take the 6 games of the 1st round and the 3 games of the 2nd round and combine them together. I’m not sure that this is meaningful. Just because RNH, Lucic, and Eberle did well against Pavelski, Thornton and Burns, does not mean they’ll do well against Getzlaf.

Now, three games is not a huge number and the Oilers are still up 2-1. However, if the match-up isn’t working, you don’t have much time to just keep trying and hope it’ll work out.

The RNH-Lucic-Eberle line that worked so well against San Jose, has not been the same against Anaheim. Let me provide evidence…

Vs San Jose

(data for this & other tables are score, venue and zone-start adjusted from corsica.hockey)

This table is sorted by CF% (corsi) but you could also sort it by xGF% (expected goals) and my point is the same. Against San Jose, the line of Eberle, Lucic, and Nugent-Hopkins lead the team in shot attempts and expected goals and they did it against Pavelski, Thornton and Burns, which is frankly amazing.

Let me explain this table. The 1st column after the name (iSF/60) is individual shots for per 60. Where as the previous table is what the team did while that player was on the ice, this table is what the individual did. So you see that Nugent-Hopkins led the team by a wide margin in shots on goal per 60. The table itself is sorted by ixG/60 (individual expected goals / 60). The last column (iSCF60) is individual scoring chances for per 60.

What you’ll see here is that Nugent-Hopkins appeared to be driving the line when it came to offensive contributions. Now, this doesn’t include passing and other plays, so it’s just one small window. However, RNH was the one getting the best chances to score, with Eberle being in the middle (though he had distrubingly few actual scoring chances for) and Lucic trailing by a reasonable margin and being closer to the bottom-6 types.

Vs Anaheim

This table is sorted by xGF%. Through the first three games against Anaheim, largely matched against Getzlaf, the RNH, Lucic, Eberle line has not been nearly as good. I suspect RNH has been boosted up slightly by his time with Draisaitl in the last game where they combined for an amazing 9 scoring chances for and only 1 against while they were on the ice together (via Natural Stat Trick).

Now, in fairness, the entire team has not been good by this metric. Only the presumptive 4th line of Slepyshev, Desharnais and Pouliot has done relatively well taking on mostly Anaheim’s bottom 6 (particularly Slepyshev). Regardless, it seems clear that the top two lines have not been getting it done against Anaheim.

I throw up this table just to balance out some of the inevitable Eberle-bashing that my previous table might contribute to. Again, individual shots, expected goals and scoring chances don’t tell the whole story. There are other things that players do to contribute, such passing plays, zone entries, puck recoveries and more. However, by goal-scoring standards, it’s Lucic that is struggling 5×5 to get any chances, not Eberle. Via corsica data, Lucic has not had a single 5×5 scoring chance against Anaheim so far and has the least shots on net out of him, RNH, and Eberle.

(Side note: Slepyshev!)

So What To Do?

The answer here isn’t clear. We have a line (Nugent-Hopkins, Eberle and Lucic) that did wonderful work through the tail end of the season and scored reasonably. In the first round of the playoffs, they were absolute money, even though the results didn’t match up. Now, against Anaheim, they are struggling, as is the McDavid line.

One option is to view these last 3 games as just a blip… a small sample of struggles in a larger sample of better performance. That’s how Sunil appears to have interpreted it and it is how I would’ve interpreted it in the regular season. I still see a reasonable argument here especially since the Oilers are still up 2 games to 1 in the series.

However, this is not the regular season. Anaheim has set up their game plan specifically to the Oilers. As a result, McDavid hasn’t been himself and the RNH line is equally struggling. It’s possible this is not a blip and the Oilers have to adjust.

I lean towards the latter.

I’d try to give Anaheim a new look to match against and spread out the scoring. There’s a risk in that, but if the top 2 lines aren’t getting it done anyways, better to give Anaheim more headaches through depth.

Slepyshev has been running against Anaheim. I don’t know if he can do it against the top lines, but I’d give it a try. Caggiula has also been okay this series and has the speed to keep up with McDavid. It’s a risk putting two rookies with McDavid, but Anaheim has been clogging up the neutral zone and keeping McDavid from moving through there. Slepyshev and Caggiula have some speed and might be able to give Anaheim some difficulty through the neutral zone while they’re focused on 97.

I like Draisaitl with Nugent-Hopkins and we saw a bit of that in the last game. Lucic played well with Draisaitl in the past as well and so I’d consider keeping them together on the line.

That drops Eberle down to the 3rd line. I’m not convinced that Eberle is actually doing more poorly than Lucic 5×5 (I think they’re both struggling relatively equally actually). However, I’d give him Maroon on the other wing to help drive that line. Maroon, of all the Oilers wingers, has most consistently shown an ability to produce away from McDavid.

Caggiula – McDavid – Slepyshev

Lucic – RNH – Draisaitl

Maroon – Desharnais – Eberle

Pouliot – Letestu – Kassian

Pros and Cons of My Suggested Lines

I fully acknowledge that first line is a dangerous suggestion. I’m suggesting we’re taking a “struggling” McDavid and saddle him with two rookies and pit him against the best shutdown unit Anaheim has to offer. It’s risky. However, the payoff could be great down the line-up. Both Slepyshev and Caggiula have speed, which they can exercise through the neutral zone. They can keep up with McDavid. I think their speed might allow for more crosses in the neutral zone that would open up room for a zone entry. Once in the zone, Slepyshev has a good shot as well for McDavid to tee up. I fully admit it’s risky, but the reward could be high.

IF that first line works, then everything falls in to place. Draisaitl and Lucic have played well together, while RNH has been golden this playoffs and has been the larger offensive contributer to his line. Draisaitl has done well in zone transitions this season and could help push the play up on this line.

The third line of Maroon, Eberle, and Desharnais would get weaker competition and could really create damage. Maroon has been one of the few Oilers to consistently drive play away from McDavid. Giving him Eberle allows him to have some skill to work with. It also is the most obvious line to slip in McDavid for a double-shift to create an instant first line if the Oilers take advantage of their last change to catch the Ducks with their bottom-6 lines out.

The fourth line also has legitimate talent. Pouliot has been doing well as part of the Desharnais and Slepyshev line.

The sum total spreads the scoring throughout the line-up, gives the team a new look, and causes Anaheim to re-shift their tactics.

In Summary

By xG, even adjusting for score, the first two lines are struggling against Anaheim at 5×5. The Oilers are still up 2 games to 1 and there’s no need to panic. However, now is not an unreasonable time to shake things up, take advantage of last change, spread out the scoring, and give Anaheim a much harder task with the line matching that Carlyle loves so much.

Thoughts?