Mediocre goalies and world-beating play; or 'Steve Mason for Vezina' (Trending Topics)

Ryan Lambert
Mediocre goalies and world-beating play; or 'Steve Mason for Vezina' (Trending Topics)

Let's just lead with the argument that no, Steve Mason probably isn't a very good goaltender. But the evidence to argue that point is mounting. 

He started the week off right by beating the Penguins — and standing on his head to do it — as another lost-cause season in Philadelphia is winding down, and hey, at least they swept their bitter rivals. In that game, Mason was magnificent in stopping 46 of 47 and building his season save percentage to a staggering .929.

As one of Mason's harshest critics over the years, Flyers fans, who have so very little to cheer about these days, were glad to have a chance to gloat that they (by virtue of having been born in the greater Philadelphia area) were right about Mason ever since he was acquired from Columbus for a third-round pick. 

Certainly, it is hard to say with a straight face that Mason has been anything but spectacular for the Flyers since he came aboard. Through the end of that Penguins game, he'd played 117 times for the Flyers, stopping 3,136 of the 3,394 shots he faced. That is a very, very good number. In fact, .924 is so good that only one goaltender in league history has a career number north of it (the still-young Tuukka Rask at .927, which seems impossibly high).

So the question is a simple one: Have Mason's critics been adequately proven wrong by this run with Philadelphia?

Flyers fans would reply with an emphatic “yes,” but the actual answer is a little muddier than that.

The problem with evaluating goaltending performance is that it's not easy. There are many ways to say that a forward or defenseman is playing well, or isn't. You can look at points, and corsi for, and corsi against, and zone starts, and quality of competition, and so on.

For goalies, we really only have a small number of data points, all of them dependent upon the work of the team in front of him to some degree or another. So conventional wisdom is that, because the Flyers are bad in front of him (they barely made the playoffs last year, and didn't come close this year), Mason must be playing exceptionally well, both relative to the rest of the league — thus the sky-high save percentage — and especially in comparison with the dismal years in Columbus (a pathetic .903 in 232 games across five seasons).

And to some extent this is true. Using War on Ice's amazing Hextally tool, we can see that Columbus actually did a pretty good job of limiting shots in the middle of the ice relative to the rest of the league with only areas of rather low consequence seeing higher-than-average levels of shot attempts from Mason's rookie year to the point at which he was traded. On these two charts, the redder the hexagons, the more shots he faced from those parts of the ice.

Now compare that with Philadelphia the last two seasons. They have one of the worst defensive corps in the league, which is why you start to see a lot of numbers that start with a “1” rather than a decimal point. They're conceding more shots than the league average just about everywhere around the net, and out near the points as well. So Mason is not only busy, but he's facing a rather high quality of shots at that increased rate. 

But what's interesting about these numbers is that even though Mason is facing more shots of higher quality, he's also (obviously) doing better with them. Here too, the one on the left is Mason in his time with Columbus and the right is Philadelphia. Here, redder the hexagons, the lower the save percentage.

So what does that tell us? That Mason wasn't seeing a ton of shots from high-percentage areas in Columbus but he was getting smoked on them. On shots from the entire slot area, and the entire area between and below the circles, he was only stopping shots at the league average at best (but congrats on beating the league average on shots from below the circle along the right wing.)

Now, in Philadelphia, he's really only getting beat from the left point and within the left side of the home plate area, and any shots from the mid-slot are getting devoured. That's a pretty good recipe for a high save percentage overall, and it explains a lot about his career's Lazarus-like resurrection.

Something very clearly changed, and very quickly. There was no gradual improvement as access to different coaching — which might have been more suited to his style and build — took effect. Mason simply showed up in Philadelphia and started posting Henrik Lundqvist numbers. It really isn't a stretch to say that he's been among the best goalies in the league over the entirety of the last two seasons.

But that's just not really evidence that he's turned his game around more or less at the drop of a hat. That's merely evidence he's had two good seasons over 110 appearances.

Now that we understand the context of those performances, we also have to think about how likely they are to be repeated. Remember, Mason has played fewer than 120 games in Philadelphia so far, compared with more than 230 in Columbus. His career save percentage, then, remains a below-average .910 and that's after playing extremely well for two full seasons.

It might be illuminating to look at Mason through the lens of a goaltender who is decidedly league average, and one who has likewise long been perceived as being awful and not worth a team's time or money. Coincidentally, both Marc-Andre Fleury (the average one) and Ondrej Pavelec (the abject one) are, like Mason, also in the midst of some of their best season ever.

The good news with Fleury is that we have plenty of information about his career to draw upon; he's been a full-time goaltender since 2005-06, and faced more than 16,500 shots in his career. That's a lot of data points, and in fact, here is his entire career in five-game rolling averages.

The thing you'll notice about Fleury is that he really hasn't gone through any stretches of prolonged and deeply low save percentages in years. It should come as no surprise, then, that the last three seasons are among the best of his career, and this suggests that the early days, when he was just as likely to go .905 as .918 or so are behind him.

Things are normalizing, and we're getting a clearer picture of this roughly average goalie as a result; he might be a little better than his .911 career number, but probably not by all that much. There's a reason most of the fluctuations in his five-game average save percentage fluctuates but never really strays too far from that .910 midpoint: That's just about what he is.

At the same time, we have far less data on Pavelec. He's only been in the league since 2007-08, and has faced about 9,800 shots. That's significantly less than Fleury. But you really only need about 200 games of information to determine what a goalie “is” — that large a sample will give you a good idea of how repeatable his save percentage is going forward — and Pavelec surpassed that mark long ago.

What you'll notice here, too, is that Pavelec has really just cut out those Marianas Trenches in his occasional performances, and occasionally posted some of the highest save percentages of his career in any given five-, six-, or 10-game stretch. But he always ends up right around that midpoint of .905 again, doesn't he? Would it therefore shock you to learn that his career average is a rather-poor .906? It wouldn't? Hmm.

So now let's think about those trends in comparison with Mason. He's played about 350 games in his career, facing about 10,000 shots. How do his numbers look over time?


The thing that should immediately jump out at you is that his numbers, unlike Fleury and Pavelec, tend to be lower as a baseline. That chart only ranges from a low of .810 to a high of .970 or so, versus .820/.830 and .990 for the other two netminders. So already Mason isn't really going on huge runs of major success apart from that one spike in the dawning of his career. Second is that he, like Pavelec, has largely cut out the drops so steep they'll give you the bends when you come back up. Look at how sharp the increase is toward the right-hand side of that graph; he's not even approaching career norms for the most part.

I tried putting all that data into one chart to illustrate how close to league average all three are in reality on any given night, and how likely that level of performance was going to be in the future. But it was a mess to look at, so I thought I'd do things a little differently. Here's each goalie's save percentage by season:

It's concerning, in a way, how much Mason just shoots up at the end there. These other goalies stay more or less within the given range of their talents, but not Mason. You need a sherpa to navigate his recent campaigns. Frankly, goaltenders just don't do that. And certainly not at Mason's age. In much the same way that you'd be skeptical of a guy jumping from an average 15 goals per season over several years early in his career to 35 per in two years with a new team, there's a lot of reason to be dubious that Mason has transformed himself this completely. The odds are that he's not this good, but rather that he had 110 really great games. 

With all that having been said, it shouldn't diminish from the fact that those 110 games were, again, really great.  The question is how long he keeps up the Icarus act — which really ought to earn him some Vezina consideration — and it comes with the acknowledgement that .929 is about as good a save percentage as you can reasonably expect even the best goalies ever to post in any given season.

Put another way, no one who isn't a Flyers fan should be ready and willing to say Mason is a great or even good goalie yet. But man, he's been great the last two years. 

Ryan Lambert is a Puck Daddy columnist. His email is here and his Twitter is here.