On this week’s episode of BSH Radio, one of the questions that came up was “who are we not getting enough from?” in terms of the team’s lineup. This is something I’ve wondered about a bit in the early-goings of this strange season — which players aren’t performing up to (or, on the opposite side of the coin, are performing above) the expectations that we’ve set for them this year? Who’s trending up or down relative to what we’re used to from them? Who’s been more, or less, productive, than they typically are?
Here, we’ll try to answer this question. One number that I like using to try and take a swing at capturing a player’s all-around production is called Game Score, a statistic created by Dom Luszczyszyn, currently of The Athletic and formerly of Hockey Graphs. There, he wrote a bit about the number when he first created it two years ago. Check there if you’d like some more details on it and have a few minutes, but we’ll try and sum up the long and short of it here:
- Game Score attempts to capture everything a skater does in a game and weights it to try and create a single number, one on a scale fairly similar to points as we know them. It takes into account an individual’s scoring (goals and assists), shots on goal, blocked shots, penalties drawn and taken, and faceoffs won and lost, as well as his team’s 5-on-5 on-ice shot and goal differential with him on the ice.
- More specifically, it is calculated with the following formula: (0.75 * G) + (0.7 * A1) + (0.55 * A2) + (0.075 * SOG) + (0.05 * BLK) + (0.15 * PD) – (0.15 * PT) + (0.01 * FOW) – (0.01 * FOL) + (0.05 * CF) – (0.05 * CA) + (0.15 * GF) – (0.15* GA). All of those statistics are across all game states except for goals for/against and Corsi for/against, which are calculated at 5-on-5 only. (There is a formula for goalies as well, but it only looks at saves and goals against, making it essentially the same as save percentage.)
- The statistic’s biggest limitations are that it essentially ignores shot quality and defensive-zone ability (beyond the ability to just prevent shot attempts, which is measured in the Corsi portion of the formula), context (players in heavy defensive roles/usage will inevitably look bad by it), and penalty killing (not that any Flyers are making contributions on the penalty kill). Even with that all acknowledged, these limitations are constant, so a year-over-year comparison for a given player should still have some validity in answering the question “how has a player’s productivity changed?”, but we’ll try and point it out where context should be mentioned.
Is it perfect? No. But there’s a lot that goes into the question of “just how much is a player accomplishing in this game?”, and this is probably about as good of a metric as any to try and answer that question in an easy-to-digest way.
Since the Flyers’ lineup right now is, essentially, the same as it was for large parts of last season, we can do a fairly clean comparison of how each player’s performance stacks up to what it was last year. We’ll compare their season-long numbers through Tuesday’s loss to Florida — based on Game Score per game played — to both their statistics through the first 18 games last year (up through their November 14, 2017 loss to the Minnesota Wild) and through their entire regular season last year. Raw data used for these calculations are courtesy of the invaluable Natural Stat Trick.
First up: the team’s top-six forwards.
Some thoughts on the above:
- It should be appreciated that Claude Giroux is putting up essentially the same level of production that he did last season, when he should have been a Hart Trophy nominee. TBD if he sustains it or even improves upon it (like he did last year, as we can see in the above), but after a somewhat slow first few weeks he’s been holding up his end of the bargain.
- Couturier’s slow start to the season has been fairly well-documented, and had it not been for some strong games last week, his dropoff compared to last season would look even worse. Perhaps more surprising is the significant dip in productivity we’re seeing from Jakub Voracek, who on the surface seems to mostly be the same player he was last year; with 16 points in 18 games, he’s only a bit off of his point-per-game pace from last year (though he’s further off from the scorching-hot pace he started last year at). The key driver of this is probably Voracek’s on-ice shot numbers; so far this season, they’ve actually been pretty poor — he’s got a 46.14 on-ice Corsi For percentage through 18 games. Voracek’s play-driving numbers have been just-OK in the past two seasons after being routinely strong through most of his Flyers career, but this would be a pretty significant step back for him if it’s something of a new normal. One possible explanation is that he’s been used in a much more defensive role this year than he usually is — only 39 percent of the faceoffs he’s been on the ice for have been in the offensive zone. Plus, it feels like he’s been better lately with Oskar Lindblom and Nolan Patrick. Worth keeping tabs on, though.
- The Reverse Voracek here, if there is one, is Travis Konecny, whose scoring numbers aren’t anything too special but whose on-ice numbers are excellent, as Jake Fahringer discussed here last week. If the puck starts going into the net for him a little more frequently, we’re going to be talking about him a lot. (Also worth noting, like with Voracek, that context may matter here: Konecny’s getting the most offensive zone starts of any player on the team.)
- There’s a gap in the data for Lindblom here because he of course wasn’t on the team in the first 18 games of last season, but for all the talk of the chemistry him and Patrick have shown, it’s pretty interesting that their numbers have looked damn near identical since his arrival to the league.
Next, the bottom-6 forwards, at least the ones who have playing time both this and last season (hence Michael Raffl’s presence on the chart):
What sticks out here:
- Those first couple of weeks of the season, in which Wayne Simmonds certainly appeared to be Back, do feel like they’re a distant memory at this point, as Simmonds’ overall productivity is closer to what it was last year (when he dealt with approximately 63 different injuries) than what it was in years past. Context is important here — Simmonds has mostly been on the third line and, more crucially, hasn’t been on the top power play for much of the season — but with his contract situation looming large, one has to think that if the Flyers are looking for more out of him if they’re going to give him an extension.
- Dale Weise looking like basically the player the Flyers thought they were getting when they signed him three summers ago has to be the biggest shock of the season so far, right?
- Scott Laughton coming in just a touch above where he was last year surprised me a bit, given how much has been made of how good he’s looked so far this year. Much like Voracek in the previous section, his good scoring numbers so far have washed over some unimpressive possession numbers (44.52 percent Corsi For as of this writing), ones which are influenced to some extent by context (he has an offensive zone start ratio of 37.23 percent). Wouldn’t be surprised if both his scoring and possession numbers regress a bit, albeit in opposite directions.
And finally, the defensemen:
- In general, the picture painted here is one of “the team’s two star defensemen haven’t been themselves to start this season, and the depth guys in the lineup have for the most part had to try and pick up the slack by way of marginal improvements”. I don’t think that’s one that’s too detached from reality, even if it is a little jarring to see just how far off Ivan Provorov and Shayne Gostisbehere have been from what we’ve come to expect from them.
- The Radko Gudas/Travis Sanheim pairing’s success in limited minutes has been a big topic of the season’s early-going, and we can see here how productive they’ve been so far. They’ve received the easiest competition and assignments of any of the team’s defensemen, but you can only play the minutes you’re on the ice for, and they’re playing it well.
- Robert Hagg’s big jump forward this year is interesting, and there’s more to it than I thought there would be given the fact that his Corsi hasn’t changed much year-over-year (45.65 percent last year, 46.01 percent so far this year). His scoring being up is probably the key driver here — he has seven points already this year after tallying nine all of last season. That said, two other things should be noted. First, he’s blocking shots at a much higher rate this year (1.83 per game) than he was last year (1.43 per game). And second, he’s been handling tougher defensive assignments, and his offensive zone start ratio has dropped to 36.5 percent (from 41.6 percent last year). I’ll be curious to see how the rest of this season goes for him, particularly because I think the scoring is going to slow down at some point.
- Not that we needed further confirmation of this, but good God, Andrew MacDonald was bad before they took him off the ice. I guess that happens when you take a third-pair-ish player and try to bring him back two weeks into a six-week injury!
- If you’re curious, Christian Folin’s Game Score per game so far this year is around 0.33, which is actually not bad at all for a defenseman who gets third-pair ice time and doesn’t really score at all. Sure, he has had some glaring visible screw-ups at times this year, but he shouldn’t come out of the lineup for MacDonald unless/until things get noticeably worse for him.