Calgary’s goaltending controversy, and trying to make sense of it (Trending Topics)

A little while back I wrote a column saying that goaltending controversies are almost always overblown by the media and that, in reality, teams usually ought to have a pretty good idea of which of their goalies is the best.

One team that is quite notably going through a goaltending controversy right now, entirely because they put themselves in this position, is the Calgary Flames.

This is a club with a lot of evidence as to which of their goaltenders gives them the best chance to win every night, and unlike most goaltending controversies, this isn’t some ginned-up story to fill column inches: By all appearances, they legitimately just don’t know which of their three (three? Three.) goalies to play. 

First and foremost, they have Jonas Hiller, a long-time NHL starter who is self-evidently at least average and sometimes a little above that. He’s in the final season of a two-year deal that pays him $4.5 million against the cap. He is the guy they should be playing most of the time because he is the one with the best career stats by far. In general, you want your goalie to be the one that’s clearly good rather than the one that is more of a mystery.

That mystery, who happened to get the start as the Flames were beaten to a pulp in the season opener by a Vancouver team that’s the definition of middling, is Karri Ramo. He’s only been back in the NHL for two seasons and never really overwhelmed anyone with his play (a little below league average both years), but when Hiller faltered a little bit the Flames switched to a platoon, and never really stopped improbably winning. Thus the roots of the controversy really began to take hold in the fertile soil of a team already believing in magic vis a vis why it had success.

Calgary’s crowded crease was further complicated by the fact that, for all Ramo’s apparent success last year, he was on an expiring UFA deal and could have walked in the summer. GM Brad Treliving instead opted to re-sign him for $3.9 million for one year, just $600,000 less than Hiller, for reasons that still aren’t all that clear except that perhaps his sub-mediocre save percentage did not actually present any tangible detriment to the team in terms of wins and losses.

[Yahoo Fantasy Hockey: Sign up for a league today]

Another $600,000 problem obfuscating the issue: Joni Ortio. He’s a rookie goaltender with just 15 games of NHL experience (pretty much all of them bad; he’s a career .899 goalie at 24 years old) who prior to 2014-15 signed a two-year deal that guaranteed him a one-way, $600,000 contract in the second year of the contract. At that time, that didn’t seem like a terrible idea because Ramo looked like he wasn’t long for the organization, which would have made Ortio — who has a decent track record both in his native Finland and the AHL — the backup.

But now there’s this mess higher up on the depth chart to deal with. As a consequence, Ortio was scratched for the season opener, and the team is carrying three goalies for at least the foreseeable future.

All of which is a long way of saying that this is a problem of their own creation, and the rumors about them trading Hiller at least made some semblance of sense because, while the other two guys clearly aren’t as good, he’s therefore the only one who would fetch much value on the market. Not that any team should trade a Hiller to clear roster space for a Ramo or Ortio, especially if they’re looking to win hockey games as the Flames apparently are, but when you paint yourself into a corner, your most desirable option isn’t going to be available to you.

So here’s the question: If you have ample evidence that Hiller is the better goalie, why do you not start him on opening night?

This question led me to wonder if there was something in the numbers I wasn’t necessarily picking up on from looking at the basic stats alone.

The first thing that occurred to me here was to check the adjusted save percentages for all three goalies. Ortio’s was garbage (.897) but Ramo and Hiller were in the same general area at least (.922 and .927, respectively). One had to wonder, though, whether that was the result of Ramo facing more high-quality shots; the adjusted numbers, as the name implies, try to control for that sort of thing, which is why the field between them comes off as being so even in the first place, but still, it stands to reason that if a goalie can post a reasonably good save percentage on high-quality chances, there’s a pretty solid chance that he will appear to be better than he actually is.

That is to say, if you take two goalies and have them face a roughly equivalent number of shots over a season and they produce similar save percentages, the guy who faces more low-quality shots is being credited with more saves that are easy to make, thus potentially inflating his numbers on dump-ins, shots from the point, and so on.

Meanwhile, if the other guy faces more shots from the low slot and around the crease, even if he allows more goals from that area, he’s also likely to be making more saves in that area, which shows up on highlight reels a lot more often. That way, even if the stats aren’t as good, the results at least appear better. This is especially true because a lot of high-quality chances are ones where you say, “Ah, he couldn’t do anything about that one.”

And indeed, Ramo faced more high-quality shots per 60 minutes at 5-on-5 than either Hiller or Ortio, and also significantly eclipsed the league average for all goaltenders that weren’t on the Flames last year. Which makes sense because the Flames weren’t exactly a shot-suppression juggernaut. When you’re facing one-plus extra shots from a dangerous part of the ice, that’s an extra chance to make the highlight reel every night. 

Which, psychologically, might start to add up.

However, Ramo’s save percentage is so much worse than Hiller’s on those high-quality chances (.853 to .844), that the number of pucks he’s stopping from there on a given night is actually worse. In fact, Ramo is also 10 basis points worse on stopping medium-quality chances (.937 versus .927), and on low-quality shots he was only negligibly better (Ramo’s .970 is a smidge better than Hiller’s .967). And interestingly, the save percentages for both on high- and medium-danger chances are significantly better than league average, while both are at least a bit behind on low-quality chances. As if we needed further evidence Calgary was a weird team last season.

So no, the reason the Flames opted for a Ramo/Hiller battery rather than just going with the obvious guy wasn’t because of stats at 5-on-5. And while that type of play makes up the vast majority of your season, there’s really only one other, far less common situation in which goalies have any sort of major influence on the game: Shorthanded time.

It is here that we start to see where Calgary’s inclinations toward Ramo probably emerge. Hiller stopped just .830 (.838 adjusted) while his team was down a man, compared with Ramo’s .861 (.850 adjusted). Now, that’s a 31-point gap between these two goalies, which is absolutely massive. But the fact that they actually get a lot closer together when the numbers are quality-adjusted probably tells us something else that’s a bit odd is going on.

It should also be noted here that Hiller’s career save percentage when shorthanded was .875, so it’s possible that, a) he had a down year, b) Calgary played horribly in front of him, c) both A and B are true, or d) None of the above, somehow.

You can throw Ortio’s numbers out entirely here because he allowed one goal from a high-danger area in a little more than 18:30 of PK time, and that was it (one goal allowed on 18 shots). But where Hiller and Ramo were concerned, the differences were stark.

As you can see, Hiller faced more than two extra high-danger shots per game, about 4.67 fewer from medium-danger areas, and nearly five fewer from well outside areas normally expected to result in goals. With all this in mind, we can safely infer that while Hiller was demonstrably worse at stopping the puck with his team shorthanded, he was also getting far less help when it comes to limiting high-quality opportunities.

So yes, that’s going to make a goaltender look less-good than he actually is. But the Flames have also made these decisions far more complicated than they ought to be, through poor roster management and misevaluation of the players in front of them.

It should come as no surprise at all, then, that Ramo’s winning percentage in the regular season last year was .611 to Hiller’s .571.

Wins are the only stat that matters when you play with this much fire. 

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

All stats via War on Ice unless otherwise stated.

MORE FROM YAHOO HOCKEY