clock menu more-arrow no yes

Filed under:

A good goalie can show you a good coach, but he can’t make you one

New, comments

Dave Hakstol’s goalies let him down. That doesn’t mean he shouldn’t have been fired.

NHL: Philadelphia Flyers at Vancouver Canucks
I guess Dave Hakstol isn’t the only one being taken out by Philadelphia goaltending.
Anne-Marie Sorvin-USA TODAY Sports

“Show me a good goalie, and I’ll show you a good coach.”

It’s a saying you’ll hear every so often in conversations about hockey, one whose origins I just spent the past 10 minutes unsuccessfully trying to find. And whoever was the first to say it was certainly on to something in how we perceive the quality of work a coach is doing.

How often do we see coaches get credit for doing an amazing job when the single biggest change made for his team was strong play in net? What do you imagine the correlation is between Jack Adams Trophy winners and strong goaltending performances? And, of course, the inverse is often true — how many coaches have lost their jobs because of losses that can mostly be traced back to their goalies not being able to stop a puck? How often do coaches whose teams are doing a lot right find themselves under fire because their goalie just can’t come up with a save when they need one?

Show me a good goalie, and I’ll show you a good coach. Show me a bad goalie, and I’ll show you a bad coach, or even an ex-coach.

It is with the latter of those two thoughts that we turn to our dear Flyers, and their recently-fired coach, Dave Hakstol. Across the past few days, as Hakstol went from “almost certainly about to be fired” to “dead man walking” to “fired”, there was a sentiment expressed in a couple of places that was at least implicitly meant as a defense of him: how can anyone win with the goaltending the Flyers have had during Hak’s tenure?

On Monday, minutes after the Hakstol firing happened, Sean Tierney, a stat-focused writer who has written over at The Athletic and Hockey-Graphs, tweeted out a thread of statistics about this Flyers team that had the recurring message of “everything would be fine if they could just get some goaltending”:

I won’t post the entire thread (feel free to click through if you want to see it), but you get the gist of his point: the Flyers have a lot of things going right for them at 5-on-5 and yet their goaltending is undermining their ability to contend, and Hakstol being gone isn’t likely to make things any better.

Later in the day, TSN’s Frank Seravalli, a long-time Flyers beat writer who is still based in Philadelphia, tweeted out the following number (and accompanying opinion) about the now-departed coach’s time with the team:

(Seravalli doesn’t explicitly mention Hakstol in that tweet, but he did so later on Monday in an article about the firing, saying that Hakstol “was also never evaluated with the benefit of proper goaltending”.)

I have some minor quibbles with certain things said here by Sean and Frank (two guys who, for the record, I like a lot and consider to be very good at what they do), but I won’t really get into those because they’re somewhat outside the scope of this piece. The overall point of each of their messages, as well as ones made by other observers both local and national, is a fair one. Because it’s probably true that if the Flyers receive league-average goaltending this season, Dave Hakstol is still the coach of this team.

And yet, if the question is “should Dave Hakstol still be the coach of the team?”, I’m unconvinced that this is an actual argument that matters in coming up with the answer.

Just how bad is it?

First, let’s try and quantify just how bad the Flyers’ goalies have been this season. You already know that it’s been real bad, but I’ll walk through it here in slightly more detail so we can come up with a baseline.

Through last Saturday’s game against Vancouver, the Flyers’ team save percentage across all game states with a goalie in net (i.e. excluding empty-net goals) was .881, according to numbers courtesy of Natural Stat Trick’s goalie page. This is, by far, the worst in the NHL this season. The difference in raw save percentage between them and 30th-place Florida (who is sitting at .890) is roughly as big as the difference between Florida and 24th-place San Jose (at just under .900). If Flyers goalies had faced the exact same number of shots that they’ve actually faced, and had stopped a league-average percent of those shots (which, so far this season, is around .907) instead of the percent that they’ve actually stopped, they’d have given up about 85 goals instead of the 109 they’ve actually allowed.

That’s a 24-goal shift, which would move the Flyers all the way from their current goal differential of minus-24 [if we exclude the team’s +1 differential in the shootout] to an even zero. Make no mistake: that is a massive swing.

A look at numbers that account for shot quality paint a similar picture. Corsica.hockey has a statistic that uses the averaged Expected Goals of each unblocked shot attempt to calculate a team’s Expected Fenwick Save Percentage (xFSv%), or the average percentage of attempts taken by the other team that you would expect to not go into the net based on the locations and types of shots taken. The Flyers rate poorly here — across all situations, their .9267 xFSv% is the fourth-worst in hockey — which is to say that their goalies have been put in tough situations. (Not surprising, if you’ve watched the Flyers attempt to kill a penalty or really play defense at any point since 2015.)

However, even acknowledging that a lot is asked of them, the Flyers’ goalies haven’t even been close to expectation. Instead, they’ve posted an actual Fenwick Save Percentage of 90.76 percent, easily the worst in hockey. The 1.92 percentage point gap between their expected and actual FSv% is, also, the biggest in hockey. And if we apply that 1.92-point gap to the 1,255 unblocked shots that they’ve faced this season, we find — once again — that had the Flyers’ goalies just performed up to what was expected given the shots they’d faced, they’d have saved 24 more goals than they actually have saved to date.

So in short: one can, pretty reasonably, make the argument that goaltending has cost the Flyers about 24 goals this season. If we tack that number on to the Flyers’ actual goal differential of minus-24, and in turn give them an even goal differential, they’d likely have around 15 or 16 wins in 31 games, rather than the 12 they currently have. They would probably be somewhere around where Pittsburgh and the Islanders currently are in the standings, on the playoff bubble and in contention for one of the East’s final playoff spots. And while that kind of team performance may have been unimpressive enough to cost Dave Hakstol his job at the end of the season, it surely wouldn’t have been enough to get him fired before Christmas.

So despite whatever other objections you may have with him, if you want to say that goaltending got Dave Hakstol fired, you’re probably not wrong. If you want to say that the team just needs a goaltender and many of their problems will go away, you’re probably not wrong.

... and?

I could talk here about the possibility that Hakstol’s long-established history of leaning heavily on one goalie until that guy is either incapable of taking the ice or of playing well has probably led to some of the struggles that these guys are facing today, and really have faced during his time in general. To run through a few speculative examples:

  • If Brian Elliott — a then-32-year-old goalie who averaged 35 starts per season in his six seasons prior to joining Philadelphia — doesn’t get asked to play in 25 of the Flyers’ 26 games between November 9 and January 6 of last season, does he maybe not break down in February, not need two core surgeries, not look terrible in the playoffs, not show up to camp this year at clearly-less-than-100 percent, and not get injured again after the season gets going? If none of that happens, do the Flyers have their best established NHL goaltender available to them right now?
  • Would Anthony Stolarz, who between April 2017 and Thanksgiving of 2018 had two surgeries on his right knee and only played in seven professional regular-season hockey games, maybe have looked a bit better in any of those losses this past week that ultimately sealed Hakstol’s fate if he wasn’t playing in his sixth, seventh, eighth, and ninth consecutive games?
  • And if we really want to jump in the time machine here, if Steve Mason isn’t asked to play in 20 of the Flyers’ 22 games between November 15 (right after a Michal Neuvirth injury suffered on November 12) and December 30, 2016 (a night on which Mason left a game with a likely hand injury, by the way), does he maybe have a bit more in the tank for the following month, a January 2017 in which Mason plays most of the team’s games and posts a brutal .883 to help take the Flyers almost all the way out of the playoff race? And if that doesn’t happen, is it maybe also not the case that Neuvirth is then asked to play in all but one of the Flyers’ first 10 games in February 2017? Does he maybe not register a .890 in those games to take the Flyers further out of contention?

I don’t know the answers to those questions. I have my guesses, if you couldn’t tell. Yes, Hakstol’s hands were often tied due to one of his goalies (usually Neuvirth) being injured. But somewhere along the line, you have to recognize that playing whoever your best guy is at a given moment every single night, even if it may help you win that night, has a good chance to be a long-term problem, particularly given the particularly injury-prone cast the Flyers have had under Hakstol’s watch. He never quite processed that, and I think that’s at least a small part of why the big-picture goalie performance under him hasn’t been pretty.

But that’s not my point.

I could also at least speculate on the impact that the Flyers’ penalty kill — which has been a mess for Hakstol’s entire tenure — has had on the goalies’ bottom-line numbers that we discussed above. Save percentages on the penalty kill are prone to a lot of volatility, and the fact that a number of goalies that have posted good 5-on-5 save percentages in Hakstol’s time have concurrently had bad numbers on the penalty kill makes me wonder just how much worse they’re possibly being made to look by a penalty kill that is dead set on hanging its goalies out to dry.

But that — even if I wanted to pin the blame for it on Hakstol and not on primary penalty kill coach Ian Laperriere, which honestly would be unfair to Hak based on what little we know about (a) any potential division of work between the two coaches as it pertains to the PK and (b) Hakstol’s power, or lack thereof, to fire Laperriere — is also not my point.

Let’s throw out any potential Hakstol or coaching-related explanation as to why the goalies have been bad and instead just assume it to be true that Hakstol’s goaltending woes were entirely an issue of the guys in net themselves simply not being good enough. Let’s go back to the number we just established, that a team that did everything else the Flyers have done this year but got a league-average save percentage from its goalies would have stopped 24 more goals than it has. Let’s give the Flyers those 24 saved goals, and say that they have an even goal differential and a basically-even win/loss record as they prepare to face the Red Wings tonight.

Again, that’s a massive difference compared to where they are in actuality. A massive difference that would make the Flyers ... an average team. Exactly what they were last year. Exactly what they were in the three full seasons of Hakstol’s tenure.

And that’s the kicker here. If this thing that we decide is totally out of his control washes out, what you’ve got left is an average team. What this group has always been under Dave Hakstol is an average team. There is little evidence to suggest that he has been able to get more than that out of them. If poor goaltending is going to sink the Flyers under Hakstol, then only great goaltending will allow them to succeed under him. All of this, on a team that the former general manager, current president, and current majority owner all have stated that they believe has the personnel to be better than average.

And if that’s the case, then really, what’s the point of hanging on?

The adage with which we began this article — show me a good goalie, and I’ll show you a good coach — is funny because, in terms of perception, it’s often true. Guys win the Jack Adams because of good goaltending and get fired within the next two years because of bad goaltending. But it obscures the reality that of course some coaches are better than others, no matter how many pucks their goalie is stopping. And a lot of coaches out there probably aren’t meaningfully better than Dave Hakstol! I’ve said this before, and I said it on Sunday when we knew that he was going to be fired: I don’t think Hakstol is a particularly bad coach. But there’s a severe dearth of evidence that suggests he’s anything more than a run-of-the-mill one, and that’s the issue at hand here.

The NHL, even moreso than with other sports leagues, churns head coaches in and out at a fairly rapid rate. Every single team has changed coaches at least once since the last lockout. Only three active bench bosses coached a game with their current team before Hakstol got hired, and the idea that the only thing keeping Hakstol from the Jon Coopers and Peter Laviolettes of the world is an actual good goaltender can probably be applied to most current NHL coaches, the vast majority of whom will also be fired within the next few years.

The Flyers will be underway with a coaching search soon, and honestly, there’s a pretty decent chance that they don’t do much better than Dave Hakstol with their next hiring. But they might, and the odds that they find someone who, at the very least, Can Win With Good Goaltending are pretty high. When we have a large sample size suggesting that he’s just going to lead his team to average results, and when taking goaltending out of the picture entirely still doesn’t paint a picture of a team with a particularly high ceiling, that’s enough of a reason to decide that it’s time for a change.

“Dave Hakstol didn’t have a chance because he didn’t have a goalie” is a factual statement that falls for the most part at the feet of the general manager that hired him. “If Dave Hakstol had a goalie, he’d look like a good coach” is a factual statement that can be stated about most head coaches in the NHL. Neither of those statements are incompatible with the idea that Dave Hakstol wasn’t an especially good coach, nor do either of them mean that he shouldn’t have been fired.