Note: This post comes to us courtesy of Twitter user @statsrespecter. Kevin Fiala has been a topic of lively debate around these parts and many others, and statsrespecter offered to take a deeper look.
Earlier this season, Peter Laviolette made the surprising decision to move talented winger Kevin Fiala off his spot on the second line.
While the demotion ultimately did not last very long, it’s become something of a go-to move when the coaching staff wants to tinker with chemistry. According to Corsica.Hockey, Fiala’s most frequent linemates other than Kyle Turris + Craig Smith are Colton Sissons + Calle Järnkrok (11 minutes played together at 5v5) and Nick Bonino + Ryan Hartman (just under 10 minutes played together at 5v5).
The results are actually somewhat favorable for Fiala-Sissons-Järnkrok; that line took 62% of the unblocked shot attempts (FF%) while it was on the ice and had 63% of the expected goal share (xGF%). Expected goals are a statistical tool designed to measure “shot quality.” This metric considers how likely a shot is to be a goal, using things like shot location, whether the shot came off the rush or a rebound, etc. I’m using Manny Perry’s model for evaluating expected goals, though there are others out there too.
The line of Fiala-Bonino-Hartman paid immediate dividends against Las Vegas on Tuesday night, their first significant time together. Fiala set up Ryan Hartman for two goals in under a minute in a game-changing sequence. Statistically, the line measured up as well: after adjusting for score and venue, they took 53% of the shots and had a whopping 70% of expected goal share.
Regardless of this encouraging return, it doesn’t seem ideal to have a player that scored 23 goals last season on the wing of a defense-oriented center like Colton Sissons in limited even-strength minutes. Similarly, Nick Bonino has been the Predators’ weakest player statistically all season and in the long term isn’t the sort of playmaker that unlocks Fiala’s scoring potential.
But what else can the coaching staff do to ignite a talented sniper that is going through a rough patch? And, more importantly, why is Fiala the easiest target whenever there’s a shake-up?
To answer this question, it’s important to determine if Fiala is actually struggling on the season, or simply experiencing a run of bad luck. Looking at Corsica, through his 12 games played Fiala has a 5v5 ixGF (individual expected goals for) of 1.65 and an actual GF of 0, which is to say that he hasn’t actually scored any goals 5v5. By looking at the delta, or difference between, actual goals and ixGF, you can make an informed deduction about whether a player is producing according to expectations, as dictated by their on-ice play.
In short: is a player getting rewarded for how well they’re playing, or are they producing in spite of themselves?
In the case of Fiala, the xGF-GF differential is 1.65—that is to say, based on his performance, he is “expected” to have 1.65 more goals than he actually does, which ranks him 19th in the league in xGF-GF differential, out of the 295 forwards with 100+ minutes of 5v5 TOI. Fiala is generating quality chances, but coming up with an unusual amount of nothing to show for it.
In terms of the rate that Fiala generates chances, he’s respectably placed in the top 40 percent of forwards using the same TOI criteria. His 0.73 ixGF/60 places him just behind names like Artemi Panarin, Jamie Benn, and Claude Giroux—and ahead of Jonathan Marchessault, Tyler Johnson and, shockingly, Connor McDavid.
Since 12 games is a small sample, what did Fiala’s stats look like over a full season? In 2017-2018, his 5v5 ixGF/60 was 0.93, good for 18th among the 199 forwards with 900+ minutes TOI. What this means is that only 17 forwards in the league that played similar minutes produced high-quality scoring chances at a higher rate than Fiala did. For comparison’s sake, Tyler Seguin was 30th in this measure, at 0.85.
So Fiala is an established threat, and he did score 17 of his 23 goals at 5v5 last season. Is this season’s xGF-GF differential an outlier, or is he the victim of chronic bad luck? Looking at last year’s Goals vs ixGF differential, Fiala actually produced just a little over his head—his expected goals number was 1.56 lower than the 17 he actually produced, so either he got a little lucky or some other small factor didn’t make it into the xG calculations.
The conclusion that can be drawn is that while Fiala is consistently a dangerous player, and while he does have cold stretches, the drought he’s experiencing currently doesn’t seem to be characteristic of his player profile. So that leaves the question—what can be done?
Personally, I don’t see anything wrong with Laviolette juggling combinations to try to spark certain players or lines. The inability for any non-JOFA line to score consistently is becoming a glaring problem. However, to single Fiala out as the source of the second line’s scoring troubles seems strange when you consider that in spite of the lack of finish, he’s creating chances at a rate consistent with a very good second liner. I would be hesitant to separate him for an extended length of time from Craig Smith in particular, since Smith is top 20 in the league in ixGF/60. If anything, it would seem to be Kyle Turris that stands out as the statistically-struggling member of that second line.
If Fiala is to dig out, the one observation that I would make is that he needs to shoot more. He has historically been a volume shooter—in 2017-18 he ranked 13th in the entire league in shots at goal per 60 minutes (iCF/60) with 17.71 shots for every 60 minutes of 5v5. This season he is ranked 49th, with a little more than two fewer shots taken per 60. Of course, he’s a career 11% shooter, so it stands to reason that even if he just keeps shooting at his current rate, eventually his luck will turn.
While the temptation to continue tinkering and making adjustments to Fiala’s game and linemates is tempting, the numbers both this season and historically suggest that the best opportunity for him to turn it around and start producing is to simply stay the course. Inevitably, goals will start coming and patience will be rewarded.
As a final note, thanks to the On the Forecheck for the opportunity to get back into writing. It’s been about four years since I last blogged (Kris Martel sent me the original analytics primer that I wrote for the Predatorial in 2013, and it was amusing to see how dated it already was). I’m hoping to bring these types of pieces to you as much as time allows and as often as OTF will allow. If you enjoy this type of thing and are interested in learning more/asking questions, feel free to follow me on Twitter—@statsrespecter. I generally review every game and break down the result from an analytics-based perspective, so if you’re into this sort of thing you can catch a lot of information that way.