COLUMBUS, Ohio -- Trying to come up with optimal line combinations in preparation of the Penguins getting completely healthy is a futile task. In the event they do have their full complement of skaters, it isn't likely to last long. It never does.
That's never stopped me -- or any of you -- from doing it anyway.
With Jason Zucker's return to the lineup nearing, possibly as early as tonight against the Wild, and Brock McGinn not far behind him, I pondered which set of 12 forwards would give the Penguins the best chance for success during the postseason.
We can talk about that on Twitter, but the exercise sent me down the rabbit hole of each individual skater's first and second half splits, and how their play is trending.
Using the Game Score model of NHL analyst Dom Luszczyszyn, I was able to get an overall view of how each skater has been performing individually during both halves.
Some of what I uncovered was obvious. Some of it was ... mildly concerning.
Before we get to that, let's take a second to understand what Game Score is.
Game Score aims to quantify a player's individual and on-ice contributions into one tidy metric for each game. The model factors goals, primary assists, secondary assists, individual expected goals, blocks, penalty differential, faceoff differential, 5-on-5 goal differential and 5-on-5 expected goal differential into its output.
Each of those categories are weighted to various degrees to keep a secondary assist king from looking as impactful as a legitimate playmaker, but the model does have limitations because it lacks context. Teammates, competition, score, zone starts, ice time and team effects all go unaccounted for. This is done intentionally to keep things from getting too noisy while remaining simple and easy to digest.
The other important thing to keep in mind is that even though special teams points are part of the equation, special teams on-ice results are not, so players like Teddy Blueger are sure to be underrated to some extent.
There's no such thing as a perfect, all-encompassing stat and Game Score doesn't change that, but I think it does a great job describing single-game performance. Furthermore, it can help shed light on a skater's consistency or streakiness.
You can read the original writeup for the model here.
The Penguins have played 68 games this season, so I split that in two to represent the first and second half. Jan. 9 is the midpoint in between their 34th and 35th games.
In doing so, I chose to include Penguins skaters that have dressed for a majority of their games, so players like Evgeni Malkin, Radim Zohorna, Zucker and Mark Friedman were not included.
Of 16 Penguins skaters that fit the criteria, only two of them have a higher average Game Score in the games after Jan. 9 than they did in the season's first half of games leading up to it.
Two.
I'll bet you can't guess both:
Your eyes do not deceive you.
Blueger and Brian Boyle are the only Penguins skaters that have graded out better in the second half of the season compared to the first half.
At face value it seems pretty bad, but Jeff Carter, Chad Ruhwedel and Mike Matheson's drops in average Game Score are all marginal. And despite somewhat heftier drops, Sidney Crosby and Bryan Rust are still comfortably in the elite category by this metric's standards.
That said, it isn't exactly encouraging that the three players with the highest average Game Score during the first half, Jake Guentzel, Kris Letang and Evan Rodrigues, have seen the three largest drop-offs during the second half.
Rodrigues has had such a fall from grace that I'm surprised he didn't land on injured reserve upon impact. Or maybe he's still free-falling.
After several highlight-reel goals and a breakout for the ages during the first half, Rodrigues has barely graded out above a replacement-level NHL player during the second half.
Having potted a mediocre two goals since Jan. 9, one of which was an empty netter, it isn't just his ability to find the back of the net that has dried up. During the first half of the season Rodrigues had the team's most dazzling on-ice metrics in terms of controlling shot quantity and quality. The second half has brought 449 minutes of Rodrigues ice time at 5-on-5 where the Penguins have been outscored 20-14 and controlled a pedestrian 49.8% of the expected goals, per Evolving Hockey.
Part of that can be attributed to Rodrigues taking nearly three fewer shot attempts per hour, but even the quality of those attempts has tapered off. His expected shooting percentage of 6.52% on unblocked shots during the second half pales in comparison to the 9.16% figure he was at in the first half.
His defensive prowess has disappeared as well. Up until Jan. 9 the Penguins were surrendering 1.91 expected goals against per hour with Rodrigues on the ice at five-on-five. That number has skyrocketed to 2.54 since then, one of the worst marks of all Penguins skaters.
It wasn't long ago that Rodrigues appeared to be the Penguins' X-factor. Now, it's fair to question his placement in the lineup.
Shifting focus to Guentzel, there's much more to be inspired by even though his average Game Score after Jan. 9 is nearly half of what it was before then.
Much like a good portion of the Penguins' roster, Guentzel has struggled to convert at the frequency we've been accustomed to. The 2.02 goals per hour he was scoring during the first half was primed to level out (only three NHL skaters are scoring that frequently on the season), but the second half has seen that figure overcorrect to 1.12 goals per hour. That's still an immensely strong rate, just not for Guentzel's standards, and certainly not when you consider he's generating 1.5 expected goals per hour during that time.
Guentzel has been a plus finisher every season of his career, meaning his goal totals exceed his expected goal totals, so there's every reason to believe his past 34 games are the result of poor percentage luck and not a sign of what's to come. Again, this is not a Rodrigues situation where the Penguins are getting absolutely nothing. With 14 games remaining, Guentzel has an outside chance to set a new career-high in goals (40 in 2018-19) and could surpass the career-best 76 points he picked up that season as well.
However, some skepticism might be in order regarding Letang and getting his average Game Score back to the point it was at in the first half. A big reason why, is that his main partner, Brian Dumoulin, has struggled mightily of late.
Through the first half, the Penguins dominated with Letang on the ice at 5-on-5. They outscored the opposition 31-20 while controlling 57.3% of the expected goals and taking 54.4% of all shot attempts. That's teetering on prime-Letang dominance.
But since Jan. 9, the Penguins have a meager +2 goal differential with Letang on the ice at 5-on-5 and have controlled 49.1% of the expected goals and shot attempts.
Not only have the Penguins seen the quality of their chances deteriorate with Letang on the ice, they've seen the quality of chances against significantly increase.
Letang needing to drag Dumoulin along for work that he's pretty clearly not suited for any longer has unquestionably hampered his effectiveness. I believe Letang has what it takes to overcome Dumoulin's shortcomings when the stakes are high, but that's not exactly something the Penguins want to have to bank on.
Letang has played well alongside Matheson this season, but putting your best offensive defensemen on the same pair together creates imbalance further down. There's no sexy solution here other than riding it out with Dumoulin and letting Letang take over as much as possible.
Rodrigues could very well be on his way out of the lineup. As for Guentzel and Letang? They're obviously not going anywhere, and the Penguins need both of them to be at the top of their game to be true contenders amongst the loaded Eastern Conference.
