Note: This is a new format that I’d like to try to make approaching analytics a little simpler and clear. Below you’ll find an infographic—there will be a chart/graph and an accompanying short description to help put the chart in perspective. If you are having trouble reading it (you should only focus on one chart or paragraph at a time), I have transcribed the text and charts below, under the graphic. You can also find an interactive version located here on Tableau.
What does Nashville lose without Filip Forsberg on the ice?
The easy answer is offense - Forsberg led the team in goal scoring until last night, despite missing six of the team’s twelve games this season. He is responsible for generating quality shots as well, where he ranks 15th in the NHL in individual expected goals per 60 minutes (per Natural Stat Trick for players with more than 50 minutes TOI at 5 on 5). The line of Forsberg, Matt Duchene and Mikael Granlund has been deadly, just as the line of Forsberg, Viktor Arvidsson and Ryan Johansen was last season. But there’s something else that’s less obvious when Forsberg is not on the ice: defense.
Below I have used three different metrics to demonstrate the value that Filip Forsberg has as a defender: first, defensive metrics for the team as a whole as well as his linemates for the 2018-19 season as well as the first twelve games of 2019-20. Second, isolated “threat” or shot rates is assessed and broken down into components - are the results on the ice defensively due to Forsberg by himself, his teammates, and/or the opponent? Last, another method of isolating impact is used to compare Forsberg to players who finished in the Top 10 of Selke voting last year - the award given to “the best defensive forward in the NHL”.
There’s no denying the offensive skill Forsberg brings to the Nashville Predators, but the results found in the accompanying graphic demonstrate that he is improving defense - all by himself at times - at a comparable rate to some of the best two-way forwards in the NHL.
Filip Forsberg’s offensive skill is well-known.
But one aspect of his game isn’t mentioned as often as it should be: his defensive ability. He may not be a likely candidate to garner Selke votes (the award for best defensive forward), but the Nashville Predators—both his linemates and, to a lesser extent, the team overall—are noticeably affected when Forsberg is not on the ice.
(Note: Team statistics are based on total team metrics for games Forsberg was in or out of the lineup. Line metrics are specifically for any time the two linemates were on the ice with or without Forsberg.)
Isolated Shot Rates Against
Last season’s first line of Forsberg, Ryan Johansen and Viktor Arvidsson (the JOFA line) was a high-powered unit that competed with the best lines in the NHL on offense—but defensively they were a mess.
However, using Micah Blake McCurdy’s 5 on 5 Isolates at HockeyViz.com, we can separate the effects a player has on opposing shot rates and danger—called “threat,” as in “threat to score a goal”—into components, compared to the league average. In this case, 0 is league-average defense, below the line is better defense than average, and above is worse.
Isolated: The impact that the player themself has on the threat of the opponent scoring just by being on the ice, independent of other factors.
Teammates: The impact of all teammates on the ice with the player—is the player getting help? Or are they being dragged down?
Opponents: The impact of the opposing players on the ice on the shot rates they generate—the JOFA line obviously saw top-line competition.
From this, we can see that Forsberg was suppressing offense (-9.2% threat to score) against top-level competition (12.1% threat to score) while his linemates were around or below league average.
The Selke Trophy: Is It Really For the Best Defensive Forward?
It has not been awarded that way.
The last 25 winners have all scored over 20 goals, and half of them scored 50 points or higher.
Josh and Luke Younggren of Evolving-Hockey.com use Regularized Adjusted Plus-Minus (RAPM) to evaluate player abilities. RAPM is similar to the above methods in that it is used to isolate the impact of a single player on the ice, separate from who they are playing with and against. While a player has very little control over which of the unblocked shot attempts taken against them on the ice turn into goals scored (goalies are voodoo), they can have an effect on the quality of those shots taken, a metric which is commonly referred to as expected goals against (xGA).
Since Selke voters do take some sort of offensive metrics in mind when voting, I used expected goal differential per 60 minutes—the difference between the quality of the shots taken minus the quality of shots against—and ranked Forsberg alongside the top ten players that received Selke votes last season.