Better Together: An Analytical Look at the JOFA Line - Part I
The trio of Ryan Johansen, Filip Forsberg, and Viktor Arvidsson have faced some ups and downs this season as injuries have reared their ugly heads. Here’s an in-depth analytical look at the numbers behind their performance.
I’m sure I was like anyone else waiting for the season to start. The offseason is a desert—a long stretch of time with no hockey-related sustenance in sight. Yes, the Predators picked up Dan Hamhuis; I remember being excited about the upgrade to the third pair. And, yes, the team did not re-sign Alexei Emelin, which was a kindness, but there wasn’t much off-season activity to keep me nourished. No major UFA signings, no major rumors…nothing.
In order to stay alive, I had to work with memories—traces, if you will, of last season to sustain me until the new season started. What kind of fantastical daydreams kept me occupied, you wonder?
I had thoughts of potential flashes of Kevin-Fiala–related brilliance, in conjunction—of course—with Craig Smith and Kyle Turris. There were times, I’ll admit, I thought about Colton Sissons in the faceoff circle. I’m even willing to admit I considered the raw power of a physical, yet skilled, fourth line. All of these ideas helped, but nothing occupied as much space in my mind as JOFA.
Yes, JOFA, the elite first line featuring Viktor Arvidsson and Filip Forsberg, centered by Ryan Johansen. This treacherous trio kept me thinking of the magic of the past, while imagining future glories beyond what I’d witnessed before. Just the thought of Viktor Arvidsson hustling down the ice on a breakaway, sprung by a perfectly placed pass from Johansen, placing the puck in the net with the precision of an incredibly fast Swedish surgeon got me through many a dreary dog-day afternoon while working road construction. Some nights I’d wake up in a cold sweat after remembering the way Filip Forsberg destroyed Samuel Girard’s dignity. My days were filled with anticipation for the on-ice wizardry I would be able to expect once the season started. I knew that JOFA would not let me down.
The season started and the Predators caught fire. JOFA was giving me all I had ever dreamed of…but then things slowed down. I knew they were failing the eye-test, but I thought for sure something had to be going on beyond just what I could see. As I’ve mentioned before, I’m not a stats guy. So, in order to dig a little deeper, I sought help from someone who is. Buckle up, because Bryan Bastin is about to take you on an incredible journey.
[Ed: Bryan Bastin makes Preds-focused data visualizations on Twitter. We’re excited to also share his work with you here.]
Examining the JOFA Line
For the purposes of this discussion, I have sorted performance of the team into three categories: “JOFA,” which is its performance when all three members of the team are on the ice; “JOFA -1,” where two of the three members are on the ice, and either the other member is on another line or out of the game entirely; and “Non-JOFA,” which is any other combination of forwards, which also includes lines with just one JOFA member.
If you are comparing the performance of one line to the other three, you’ll have TOI differences (JOFA gets an average of 16-17 minutes of even-strength ice time, meaning the rest of the team combined gets about 33-36). We control for this by expressing many of the statistics in terms of “per 60 minutes,” projecting those totals to a rate if they had sustained that for a full 60 minutes. This evens out the totals you get—if the JOFA line manages to take 10 shots in 10 minutes, and the rest of the team manages 20 shots in 40 minutes, intuitively you’d think the rest of the team performed better, but those rates actually end up as 60 shots/60 minutes and 30 shots/60 minutes respectively: the JOFA line is taking twice as many shots per minute as the rest of the team is.
What we’re going to look at here are Corsi for (CF), expected goals (xG), and PDO. You can catch up on Corsi and PDO with this Sportsnet primer if you’re not familiar, though we’re looking at differential instead of share. I’ve used Evolving Hockey’s model for expected goals—you can read more about how they calculate them here.
Early Season (Beginning of the Season to Early November)
The JOFA line came out to a hot start, and the record matched: the Predators started off with a blazing 10-3-0 record in games where the JOFA line was intact. The line was getting more quantity (+11 CF/60), better quality (+1.54 expected goals/60), and the results to show for it (+2.29 goals/60).
The rest of the team over that span were much more pedestrian: -1.04 CF/60, +0.1 xG/60 and +0.6 goals/60. They were getting outshot by a small margin, the quality was nearly equal, and yet they were outscoring the opponent by half a goal per hour. When the JOFA line was reduced by one (with that winger usually replaced by Kevin Fiala), we saw a good goal differential, but negative quality (-1.18 xG/60) and very negative Corsi differential (-11.89 CF/60). In other words, these combinations only amounted to around 30 minutes of ice time, but it seems they were more lucky than they were good.
Clearly, the JOFA line was carrying the team and doing so with great success, but was it sustainable? Looking at the rolling PDO average over 3 games, the line was shooting anywhere from 18-33%—which, when league average ranges from 8%-11%, is easy to see was unsustainable. However, it was necessary, as goaltending with the JOFA line on ice was relatively weak, hovering around the mid 80% range. These two factors led to the line having a PDO above 100 for nearly the entire month, averaging around 103.
Joey Alone: Arvidsson and Forsberg go on IR (Early November to Late December)
Arvidsson went on IR for all of November, with the exception of November 10th, and did not return for good until after Christmas. Forsberg left the lineup in December.
Without the JOFA line, the team managed a 12-11-1 record (7-6 in November missing Arvidsson, 5-5-1 in December without Forsberg and Arvidsson), and this shows, as the goal differential rate was zero throughout. The Predators attempted to make up for their loss up top by shooting, and doing it a lot, as evidenced by Forsberg and Johansen’s +9.66 CF/60 while the rest of the team registered a +6.19 CF/60. When Forsberg went on IR, the rest of the team picked up the pace at a +9.64 CF/60, only dipping +2 CF less per hour than the JOFA line in October.
While the team managed to maintain quantity, the real hope was that they could maintain quality. Unfortunately, that did not happen. Forsberg and Johansen, in November, barely treaded water, with a tiny +0.11 xG differential rate (compared to JOFA’s +1.54 in Oct). The rest of the team managed a dismal -0.08 xG diff/60 in November and +0.37 xG diff/60 in December. The team was keeping up with the quality of shots of their opponents, but they were doing so by firing a LOT more low-quality shots (usually from the point).
In short, without an intact JOFA line, the team increased the volume of their shots, but only managed a nearly even xG differential and a flat goal differential, which together managed 25 points in the standings out of a possible 48. Without one or two members of the JOFA line, the team was able to keep the ship afloat, but just barely.
Would the ship get back on course as the JOFA line reunited when the calendar switched to 2019? Stay tuned for our next installment to find out!