Justin’s 5th Annual Preseason Top 30 Goaltenders List (10-1)

Welcome to the final edition of this year’s Top 30 goaltenders. We’ve been through twenty of the top keepers in all the land over the last few weeks, but here is the moment we’ve all been waiting for. Just in case anyone missed the first two entries, they can be found here and here. The first one covers all the introductory info and criteria, so make sure to check that out if you haven’t already. Get your tar, feathers and NHL Network/ESPN/TSN Top 10 lists ready to waive in my face….here are my Top 10…

  1. Jonathan Quick- Los Angeles Kings (last year’s ranking: 4)

latimes.com- Quick

At age 30, Quick is what he is; a ridiculously gifted athlete who has the uncanny ability to make highly difficult shots look easy and easy shots look highly difficult. My favorite way to describe Quick is that he gets drunk on his own mobility. He knows he has elite tools and becomes overly reliant on them, rather than staying within him game in a more disciplined fashion, which, in my opinion, would be much more effective and give the Connecticut native far greater consistency.

That said, he can still put together streaks when he looks impossible to score on. He has a penchant for doing it in the playoffs, which is why he boasts two Stanley Cup rings. Certainly not the most conservative or reliable way to address your goaltending needs, but the talent/potential for dominance is absolutely undeniable.

  1. Corey Crawford- Chicago Blackhawks (last year’s ranking: 10)
(Photo by Jonathan Daniel/Getty Images)-Crawford
(Photo by Jonathan Daniel/Getty Images)-Crawford

Crawford is more or less the opposite of Quick. Having not been gifted with special athleticism or mobility, Crawford is a success story of adjustments. He has successfully compensated for a lack of elite tools with elite positional instincts and an incredibly simple, consistent technical game plan. There is no wasted talented with Crawford.

Over the past year or so, he has improved his flexibility and ice level game. His sliding mobility is far superior than it has been in recent past and he has widened his stance up to complement his deeper positional set up. This is likely a result of greater hip flexibility. Think of Crawford like the ultimate systems quarterback. He has tailored his style to the strengths of Chicago’s systems, and it has been a highly successful marriage.

  1. Roberto Luongo- Florida Panthers (last year’s ranking: 9)
Credit: Sergei Belski-USA TODAY Sports- Luongo
Credit: Sergei Belski-USA TODAY Sports- Luongo

What a story Luongo has been since leaving the Canucks two years ago. Since arriving in the Florida, Luongo has cleaned up his technical game considerably and has morphed into the consistent, battle tested leader in the crease for a young, entertaining Panther’s team. This is to mention nothing of his NHL best twitter account.

On the actual goaltending side, Lu has made significant adjustments in his positional discipline and has improved recovery and second save technique. In the past, he used to be caught out of position frequently on rebounds and lateral plays. He has remedied that be being less reliant on athletic saves and allowing his positioning and instinctual blocking style to really take over and allow himself more time to recover when the defense falters. Think of Luongo as a fine wine, just getting better with age.

  1. Steve Mason- Philadelphia Flyers (last year’s ranking: 14)

www.philly.com- Mason

This is going to be one of, if not the most controversial ranking on this list. Last year, I ranked Mason 14th on the back of a very strong season for the Flyers. I have always loved his talent, but after the performance his last few year’s in Columbus, I was a bit skeptical that he had fully turn the corner. Well, Mason has been elite for two full seasons now behind a very poor blue line in Philadelphia. Just in case you were wondering, that wasn’t hyperbole. Mason has been the definition of elite.

From a scouting standpoint, he has incorporated greater consistency into his post-integration techniques, like the VH and Reverse-VH, which has been an issue in the past. He has been a “whole is greater than the sum of his parts” on the tools side, but he has far greater poise and consistency than the showed anytime outside of his rookie season with the Jackets.

On the statistical side, he has been an advanced stats darling. His rate stats are nothing special, but that is more indicative of his team around him. No matter what metric you look at, he comes up aces. High-danger shots? Elite. AdjGSAA/60? Elite. Shot distance data? Elite. All for a terrible team. If you are a standard GAA/Sv% person, we are just never going to agree on Mason, but if you look beyond those stats, you see a very good goaltender in his prime years who would be the man in net for almost every team in the league.

  1. Ben Bishop- Tampa Bay Lightning (last year’s ranking: 11)
Credit: Matt Kartozian-USA TODAY Sports- Bishop
Credit: Matt Kartozian-USA TODAY Sports- Bishop

Bishop is an anomaly. It is extremely difficult for goaltenders his size to be as effective as he is. I have a mental block about Bishop because he still utilizes many dated techniques and hasn’t fully embraced modern mobility tools that could benefit his game greatly. However, his positional instincts, coupled with his size, create a goaltender that is almost impossible to score on when he is on his game.

Because Bishop’s game is really about maximizing blocking surface and maximizing positional efficiency, his appearance is highly technical out on the ice. He is a battler to some extent, but there seems to be an emotional disconnect when he is at his best. Not that it has been interpreted as a lack of effort or commitment, but I would speculate that some coaches view it as less of a personal interest in keeping the puck out of the net, like say Jonathan Quick has. Curious digression aside, Bishop is a very talented goaltender who should put up some very strong numbers for Tampa Bay this season.

  1. Tuukka Rask- Boston Bruins (last year’s ranking: 4)


Rask took a bit of a step back this season, partially because the Bruins, as a whole, took a big step back this season. From a scouting standpoint, it seems that Rask has fallen in love with/become a little too reliant on the reverse-VH technique (see photo). This has limited his ability to compensate for the declining net front coverage of the Bruins’ defense, which, I would conclude has eroded his high-danger save percentage.

This combined with his increased focus on integrating more athleticism into his style has taken a toll on his positional game/discipline. For this coming season, I would advise Rask to get back to basics and the core strengths of his game, as he certainly has the talent to remain in the upper echelon on the league for the foreseeable future.

  1. Braden Holtby- Washington Capitals (last years’ ranking: 5)

washingtonpost.com- Holtby

Our regular readers here know that I love me some Braden Holtby. The reigning Vezina winner had an absolutely ludicrous first half before slowing down somewhat in the spring. He still had a magnificent season overall and at age 26, is starting to move into those precious prime years. I believe that the biggest knock on Holtby last season was the lack of a reliable back up, which saw Holtby make 66 (!) starts for the Caps. No wonder he wore down a bit.

At this point, Holby is basically Quick 2.0, as he has all the elite carrying tools that made me enamored with his game way back in 2012. The difference is that Holtby currently possesses a far greater sense of discipline within his style and prioritizes positional efficiency to a much greater degree. He has shown that he can up his game in the post season and is a high-level puck handler, to boot. There are very few goaltenders out there providing the combination of talents that Holby brings to the table and I will only assume he will continue to rise up this list.

  1. Cory Schneider- New Jersey Devils (last year’s ranking: 7)
(Photo by Rich Graessle/Icon Sportswire)- Schneider
(Photo by Rich Graessle/Icon Sportswire)- Schneider

Schneider may the most criminally underrated goaltender in the NHL. Playing on a horrific NJ Devils team, Schneider has remained in the top tier of goaltenders working today. His positional efficiency is off the charts and makes the most out of every technique.

He is not flashy or particularly impressive in his execution, but it is highly effective and he makes many high difficulty shots look routine. If I had my pick right now, he would be the man in net for Team USA at the World Cup, which could finally show the world just how very talented Schneider is.

  1. Carey Price- Montreal Canadiens (last year’s ranking: 1)
(Photo by John Crouch/Icon Sportswire)- Price
(Photo by John Crouch/Icon Sportswire)- Price

The only thing keeping Price out of the top spot for the second year in a row are the recurring injuries that have sidelined him for the past year. Since this exercise is a one-year recommendation, it makes it tough for Price to repeat in the light of potential for re-aggravation and more missed time. Injuries notwithstanding, Price is still unfairly talented, with elite carrying tools across the board and a level of efficiency and poise rarely seen. He is truly a generational talent.

His play internationally is when he is the most entertaining to watch, as the Habs defense is fairly atrocious and he is routinely hung out to dry during the regular season. However, he always finds a way to buoy them up, which is what an elite goaltender does.

  1. Henrik Lundqvist- New York Rangers (last year’s rank: 2)

The King

It seems only fitting that Hank reclaims his throne on the 5th anniversary of this list.   Despite the absurd narrative that Hank had a “down year” last season (read this article that thoroughly dispenses with that notion), Henrik’s performance has never been better. Other lists have had Hank as far down as tenth, which is far more indicative of team performance than Lundqvist’s individual contributions to a very mediocre squad from 2015-2016.

Call me a homer if you want, but Lundqvist is still the best in the business and the statistics bear that out. While they do not paint a perfect picture, they are a testament to how spoiled we Rangers fans have become watching this level of excellence inhabit the Garden crease for the past decade. His consistency, durability and elite production have cemented his as both a future Hall of Famer and the top spot on this year’s list.

This concludes our rankings for this year. Thank you for reading and being a part of our community here at BSB. As always, this has been a very enjoyable exercise for me to put together, so if you have any feedback on format, analysis, criteria or other suggestions for next year’s list, feel free to drop me a line either in the comments, on twitter or via email. Have a great weekend, everyone!

"Justin's 5th Annual Preseason Top 30 Goaltenders List (10-1)", 4 out of 5 based on 9 ratings.

64 thoughts on “Justin’s 5th Annual Preseason Top 30 Goaltenders List (10-1)

  • Sep 2, 2016 at 8:02 am

    Sorry, Lundqvist just cannot be number 1 after last year’s performance. It may have been just an off year for him, but it does count against him.

    • Sep 2, 2016 at 8:55 am

      Sally, what are your thoughts to the article that Justin linked?

      Hint: click the blue “read this article” above.

      On what metrics, or standards, do you claim that Lundqvist had an ‘off year’? BY many counts, the team around him floundered, giving up more dangerous shots per game, yet he was able to keep his standard stats (SV%, GAA) around career averages. Good read if you have the time.

      • Sep 3, 2016 at 2:17 pm

        Yea, Sally, what he said^

      • Sep 4, 2016 at 9:56 pm

        Hank had one of the best seasons of his career last year. Couple the average distance of ES shots faced (about 30 ft) and his ES Sv% .945 with a horrendous defense and PK in front of him. Makes it much easier to understand why Hank had such an Un-Hank like campeign. Gorton fixed 2 of the 3 issues noted above and I whole heartedly believe he will address the defense before the season starts. Either before the WCH2016 or before the regular season begins. My best guess is right before or after the WCH while there is a freeze on trades, contracts, and extensions.

    • Sep 3, 2016 at 8:28 am

      Got to agree. No matter the metrics, the eye test puts Hank below Price, above Rask and somewhere around Scheider and Holtby.

      People have been saying for years that Hank is the defensive savior of this team. Even when Tortorella had them all packing in and suppressing shots instead of playing offense. I don’t buy it.

      Last year he was under siege, but look what happened to Montreal when Price went down. What happened to the Rangers when Hank went down year before last? The team had an amazing run to the President’s Trophy in front of its backup.

      Sorry, let the thumbs downs fly but this obsession with Hank being the king of all goalies is homerism.

      • Sep 3, 2016 at 2:18 pm

        you do have a compelling argument!

    • Sep 4, 2016 at 10:34 am

      That was also the worst 6 defenseman he has ever had to play behind. Considering how bad the defense was, last season was one of the best of his career. He led the NHL in even strength Sv% and had to face the most difficult shots in the NHL at even strength. On average the average distance of shots faced was lower (closer) than any goalie in the NHL.
      The stats don’t lie. Just because the Rangers couldn’t keep the puck out of the net doesn’t mean Hank had a bad season.

  • Sep 2, 2016 at 8:10 am

    Hank is number 1 hopefully he will have a bounce back year

  • Sep 2, 2016 at 8:52 am

    Gotta disagree with the top 2 choices–Lundqvist was #5 at best in the NHL last year, and Price didn’t play enough for the Habs before being injured again. Schneider, Holtby, and Bishop were all better last season, in my opinion.

    Rask is a poor choice at #5 as well–wouldn’t have put him in as a top 10 netminder from last season. Luongo should’ve definitely been higher on the list as well.

    • Sep 2, 2016 at 11:18 am

      You are certainly entitled to your opinion, however if you read the first post of the series, the criteria is how I would advise a team to pursue a goaltending solution for this coming season, not ranking performance from last season.

      • Sep 2, 2016 at 2:27 pm

        Tuuka is not the goalie he used to be. Martin Jones and MA Fleury could be considered better players at this point, and I’m not sure I’d put Rask ahead of Mrazek in Detroit.

    • Sep 2, 2016 at 11:39 am

      If you were rating centers thirty years ago and Wayne Gretzky missed an entire year, would you have downgraded him?

      Rating Carey Price is tricky, but basically what we know is what we knew a year ago — except that we know that his team (unlike Hank’s incidentally) actually does stink without him.

  • Sep 2, 2016 at 8:55 am

    Any good board needs objectivity with a dose of home-team prejudice. This site tries to use statistics in many cases to bolster it’s position and make cases to support scenarios for and against certain situations – I get that. What I don’t get is how we can totally ignore something like GAA in the NHL, where Lindqvist wasn’t even in the top 10. Is Save % also a worthless stat? Don’t get me wrong, he is a great goalie, probably top 10 in the league, but how you can name him number 1 and ignore a stat like GAA and Save % just makes the the list lose credibility IMO.

    Where are the stat guys who love Corsi? Do you love Corsi but this GAA and Save % are useless stats?

    • Sep 2, 2016 at 9:39 am

      Apologies in advance for the long post …

      Sal, all stats aren’t created equal. Much like using wins and ERA in baseball is now an antiquated way to judge a pitcher’s performance, GAA and, to a lesser extent, SV% are also become antiquated thanks to the development of far superior stats.

      The idea behind this next generation of statistics is to try and remove all the surrounding noise that can make some stats tell the wrong story. If a goalie has a poor defensive team in front of him, of course his GAA is going to be high. There are only so many PK saves you can make and so many shots from dangerous areas you save. Eventually, a goalie gets overwhelmed because his teammates aren’t holding up their end of the bargain.

      Now that there are statheads diligently tracking shot location and shot quality, we can visualize and categorize the quality of the shots against a goalie faces throughout the season. Since we now have that ability to look at relative quality it makes looking at raw numbers (Shots, Saves, Goals) pointless. Why should a goalie be penalized statistically for not stopping, say, a 2 on 0? It wasn’t the goalie’s fault his team broke down and two opponents were allowed to break in alone. Of course, if the goalie manages to stop a 2 on 0, we can and should still give him credit for that. That’s why it’s important to look at SV% on shots from quality scoring chances. That’s where Hank shines, and that’s where a goalie can really make a difference.

      Any kid from the ECHL or Juniors can come in to an NHL game now and stop open shots from 50 feet away. There’s no point in giving an NHL goalie credit for those saves, he’s expected to make those 99.99% of the time. Goaltending technique and equipment have come so far that a goalie wiffing from that distance is like a place kicker missing a PAT. It happens, but it’s no longer statistically relevant. Of course, if it happens in a game, Mason last year in the playoffs, Richter in the 1992 playoffs against the Pens, it can bury a team. But those are flukes.

      I hope that helped shine a light on why someone might prefer Corsi but would think GAA and SV% and +/- aren’t worth looking at.

      • Sep 2, 2016 at 9:51 am

        That may all be true, but not even in the top 10 in either category? That is a lot of 2 on 0’s and a lot of shots in front. Hank is a great goalie, but he needs to at least be in the top 10 even with a porous defense. This is like the kid who always blames the other kid or the teacher. When can we actually say it may be Hank who has fallen off a bit?

        • Sep 2, 2016 at 10:05 am

          Right, and intuitively, what you are saying makes perfect sense. I have two counterpoints for that.

          One, the Rangers PK was dreadful, so that will depress Hank’s GAA and SV% considerably. The old hockey adage is that your goalie has to be your best PK’er, but I witnessed a disgusting amount of backdoor slam dunks and tap-ins that I can’t blame Hank for. It kind of blows that old timey folksy wisdom out of the water. Sometimes a bad PK is just bad, and the Rangers had a bad PK last year.

          Two, I remember reading this summer that Hank faced the highest number of shots against from dangerous locations in the NHL, and if I remember correctly it was by a significant margin. Factor that in with his 14th ranked SV% of .920 and that’s rather impressive. Gary’s comment below about the average shot Hank faced last year coming from 3 feet closer in than the average NHL goalie speaks to that.

          • Sep 2, 2016 at 11:20 am

            GAA and Save % include so much noise that they are useless for ascertaining talent level. They do have some merit as measures of performance, but there is still too much noise to rely on them too heavily.

            This list is intended to be an amalgamation of performance and talent. My advice to a hypothetical team. Lundqvist’s underlying performance was far superior to his rate stats, which fuels the narrative of a “down year’, which isn’t remotely true.

          • Sep 2, 2016 at 11:49 am

            JFTR, the Ranger PK was actually fine when Raanta was in goal. It was only dreadful when Hank was in goal.

            From what I saw, it didn’t look like Hank’s fault, but what we are doing is picking the numbers we like.

            • Sep 2, 2016 at 12:09 pm

              Right, but when you look at the backup vs. the starter you have to take into account that the backup is likely playing against inferior teams in most games.

              I’m not saying that you’re wrong, just that there is a chance the numbers you are breaking out are misleading you.

              • Sep 2, 2016 at 12:31 pm

                I answered this previously. There was not much of a difference in the offensive capabilities of the opponents. I sorted through the opponents and I estimated that had Raanta faced the same opposition as Hank, he would have yielded one additional goal over the course of his 19 games. Even if it was a PK goal (less than 50% likely), it would scarce have changed the numbers.

              • Sep 2, 2016 at 12:48 pm

                Sorry Ray, I missed that reply last time.

                Just looking at Raanta’s game log I see a lot of non-playoff teams on there. Yes, it’s not the most scientific approach, but I don’t think it’s fair to say they faced teams with the same offensive capabilities.

              • Sep 2, 2016 at 1:11 pm

                What I did was to look at Raanta’s 19 decisions and compute the average goals the teams scored over the course of the season. This group was about three goals IIRC below the league average.

                Sometimes just looking at opponents can be misleading. Raanta faced Calgary twice. Calgary was actually an above average offensive team – their defense stunk obviously. And Raanta faced some good foes, like the Caps.

                So your complaint is not without merit, but the opponent discrepancy is quite small (although if one looks at save percentage, an adjustment triples Hank’s one point advantage.)

            • Sep 2, 2016 at 12:13 pm

              Where can i find this stat, I would like to review. The eye test says that the NYR PK was horrible and it looked more like a failed system than anything else.

              • Sep 2, 2016 at 12:18 pm

                The system is fine, the execution was awful. The Rangers play a fairly conventional PK. The breakdown was the poor reads and slow reactions from the two men down low.

                No system in the world can help if the Dmen are too slow in reading and making their defensive slides.

              • Sep 2, 2016 at 12:27 pm

                I can’t remember exactly – think it was war on ice — but anyway goalie save percentages are differentiated by things like even strength, PK. Hank’s PK save percentage was near the bottom while Raanta’s was about average. IIRC, the difference was more than 50 points.

    • Sep 2, 2016 at 9:51 am

      GAA is pretty bad for evaluating a goaltender because it is more based on the team than the goalie. Save % is better, but doesn’t take into account shot quality. Adjusted goals against above average per 60 uses takes the distance of the shot into account and how likely an average NHL goalie is to save that shot. For example, Lundquist faced an average shot distance of 31 feet, compared to the league average of about 34 feet. Thus his save percentage is actually very good considering this was 3rd worst in the NHL.

      • Sep 2, 2016 at 10:02 am

        Good info. Thanks Gary.

        Advanced goalie stats are my next frontier to delve through. Whats the best sites to get the numbers from?

      • Sep 2, 2016 at 11:54 am

        This isn’t even true, I’ll bet. The distances are likely measured to the net and since Hank plays deeper, he is actually further from the shooter than another goalie with the same distances.

        • Sep 2, 2016 at 12:53 pm

          It’s measured to the goal line, which remains static.

        • Sep 2, 2016 at 12:53 pm

          Please explain how this would make any appreciable difference. The shot distance and type, along with whether the goaltender saves it or not, are the only relevant factors in that analysis.

          • Sep 2, 2016 at 1:18 pm

            Shot distance is measured to the goal line, as Dave notes. However, the relevant stat is actually distance to the goaltender since that tells you how much time he has to react. A tender standing three feet from the net has 31 feet to react to a 34 foot shot. A tender on the goal line has 31 feet to react to a 31 foot shot. So the shorter shot poses no problem. Now I don’t think actual examples are this extreme, but Hank has more time on closer shots than other tenders.

            • Sep 2, 2016 at 1:54 pm

              Can’t side with you here, Ray. The closer to the goal line, the more time to react, but less angle to cut, thus showing more net.

              Ultimately all that is up to the goalie. Shot distance from the goal line is static, so we can gauge how goalies fare against various types of shots irregardless of their style/position. The goalie can play the shot however and wherever he wants in his crease, and in the end the numbers will tell us who does it best. We’re not trying to gauge which goalie has more or less reaction time.

      • Sep 2, 2016 at 1:01 pm

        Is the difference between 31 ft & 34 ft statistically significant?

        • Sep 2, 2016 at 1:20 pm

          surely. Over the course of the season, the difference adds up to a mile.

          • Sep 2, 2016 at 9:33 pm

            So show me the statistical significance test that proves that.

            • Sep 4, 2016 at 10:12 am

              I doubt that work has been done.

              But for comparison’s sake, move the pitcher’s mound up three feet. See what happens.

        • Sep 2, 2016 at 1:55 pm


    • Sep 2, 2016 at 11:44 am

      For the record, save percentage is the companion to Corsi. By this standard, GAA is simply a variant of save percentage which penalizes goalies who face more shots and should be ignored.

      But I am still in the camp that save percentage is the gold standard.

      • Sep 2, 2016 at 12:00 pm

        Ray, how do you account for the massive amount of noise contained within save percentage when comparing goaltenders?

        • Sep 2, 2016 at 12:50 pm

          With so many shots, there is far less noise over the course of a season than you think. Truthfully, there is far more noise in any possible alternative because you have more categories and no additional samples.

          The real problem is not noise – it is the alignment of deceptive information. This is a problem for save percentage. Most would agree I think that Cam Talbot faces tougher shots than Corey Crawford. And the new stats try to counter this. So far so good.

          According to the new stats, the Ranger defense (which heretofore was considered pretty good – and still is faulted mainly for not getting the puck up ice) is allowing an incredible number of very tough shots and Hank is stopping those tough shots. Well, that means the metrics are arbitrarily deciding that the Ranger system is bad even though the goals aren’t going in.

          One can’t use this, but the best measure of a tough shot is whether or not it goes in the net. This suggests the shots are not so tough. In fact, we absolutely know that the Ranger defense does try to give Hank shots that he likes to handle and that this mix is not the same as the mix they give his backups.

          Hank faces tougher shots than Raanta (by the metrics) because he wants to – because his definition of a tough shot is not THE definition.

          So the new stats are biased by goalie preferences and defensive styles.

          • Sep 2, 2016 at 12:58 pm

            Ray, not sure I agree with your assessment here. How can we rely on a measure of shot quality strictly by whether the puck goes into the net? The quality of the goaltender in the way makes the criteria a moving target. If Hank stops it, but Ben Scrivens doesn’t, it is a tough shot?

            For objective shot quality data, I would proffer the following criteria weighted and measured:

            1. Shot distance
            2. Shot type (forehand, backhand, deflection, etc)
            3. Lateral distance and angle, if any on one-timers (royal road analysis)
            4. Rebound, sequence of shots
            5. Shot speed
            6. Zone entry data

            This is an awful lot to put into it, but shot quality has a ton of moving parts.

            • Sep 2, 2016 at 1:05 pm

              Man I can’t wait for the new tracking systems to be perfected. I think it’s really go to turn hockey on it’s head, in a way that’s going to make the sport even better.

            • Sep 2, 2016 at 1:39 pm

              We can’t measure shot quality that way if we want to rate tenders of course, but it serves as a check on our criteria. A tender doing ridiculously well against our tough shots is suspicious.

              Anyway, the problem with your criteria is that I could probably add ten more things that you have left out. And so any of them are defense-dependent. Which means that ignoring my extra criteria injects bias.

              I have a favorite example of a shot where Marc Staal positioned himself so that the shooter had to shoot to Marc’s left (or the shot would be blocked). He left Hank with a clear view and the knowledge that the shooter only had half a net to shoot at. In truth, Staal did all of the work, but your criteria doesn’t take it into account.

              There is the extent to which defenders harass the shooter and the extent to which the goaltender is screened. There is also the level of distraction faced by the tender. Is the goaile ready for the shot – or is he thinking about the player in the crease that might be passed to?

              All the objective standards I have seen look at the actual shots and not the environment in which the shot takes place – and I would argue that the latter is more critical.

              This is a silly example, because it rarely happens, but I think all criteria regard breakaways and two on zeros as the same — maximally difficult situations. However, goalies can stop breakaways. Stopping a two on zero is hopeless unless the opponents mess up.

              • Sep 2, 2016 at 2:22 pm

                The problem I have with that conceptual analysis is that a goaltender cannot be separated from the environment he finds himself in. The only thing we can do is try to isolate what criteria factor in most heavily into the independent factors the goaltenders themselves can control (reaction time, positioning, rebound control, etc.).

                There is always going to be some statistical noise from the other five players on the ice. Even if we simply come up with a reasonable standard deviation.

                You are absolutely right, there are many more variables we can add. The more information, the more refined the analysis and the better, more reliable statistic we create.

              • Sep 2, 2016 at 3:29 pm

                There are a number of points we should be able to agree on. One is that it is not a level playing field and that some tenders face a tougher mix of shots than others. Another is that if you choose the criteria for judging tough shots incorrectly, it may be valueless – or even worse, tell us things that are wrong.

                Even something as benign as using save% vs. GAA rewards a goalie like Lundqvist who never uses a poke-check to thwart a shot.

                Where we disagree is on whether the new stats clarify or mislead. My instincts just say they leave too much out and paint a picture that isn’t believable.

          • Sep 2, 2016 at 1:09 pm

            How can one say this is inane without coming across as condescending?

            One, there is still a TON of noise in a season’s worth of shots, whether or not you break it down into danger zones. Lundqvist faced the most shots on goal in the league across all situations this season and it was still fewer than 2000 shots. We already know that we can’t really begin to feel confident about save percentage until we cross the 5000-6000 threshold, so to say that there isn’t noise is silly.

            Two, if you’re only looking at raw save percentages, you leave the potential wide open for results to be skewed by massively differing volumes of shot location and shot type. All shots are not equal. This is not even touching on the fact that raw save percentage includes the PK and PP, which are heavily systems-based and have varied effects from team-to-team and goaltender-to-goaltender.

            Is there noise in breaking down by danger zones? Yes. But there is still significant noise when you don’t, and the insight gained by breaking down is much more valuable than the small increase in sample size from leaving it untouched.

    • Sep 3, 2016 at 2:21 pm

      Sal! nice job with even more compelling argument! I love this place!

  • Sep 2, 2016 at 9:43 am

    Great list as always Justin! This is one of my most anticipated hockey posts every year. I love hearing a goalie’s perspective on goalkeeping because they are really the only ones that understand it.

    I had a question about Rask and his #5 ranking. You make him sound sloppy and undisciplined, the way he chooses not to adapt to the team that is crumbling around him. Is that coaching or stubbornness? And if it’s stubbornness wouldn’t that really warrant a sharp drop in the rankings? It sounds like his technique is all wrong and effectively making him a poor NHL ‘keeper. Am I being too harsh?

    • Sep 2, 2016 at 11:22 am

      Thanks, Chris. Much appreciated.

      As to Rask, I don’t think it’s necessary either coaching or stubbornness. I think he, like most human beings, default to what has worked in the past. The Bruins defense used to be a major strength, which allowed him to be a little less disciplined and seek out that athletic save.

      Now, with the defense more or a weakness, he needs to be able to cover more offensive options and hedge a little more, taking reaction more to the background and positioning more to the foreground. It will definitely take an adjustment on his part, but he is fully capable of making it.

      His talent level is still sky high, he just needs to understand that he needs to reign it in a little bit. Nothing a good goalie coach cannot help correct.

  • Sep 2, 2016 at 10:36 am

    In some other news – Chris Drury named Asst. GM. – I guess snagging Vesey helped him climb the ladder pretty quickly. Best of luck Chris!

    • Sep 2, 2016 at 1:08 pm

      He landed the big FA fish. That definitely deserves a promotion!

  • Sep 2, 2016 at 12:20 pm


    Let me start with a thank you. This is a very nice list and we really appreciate your perspective. A lot goes into this I know and you’ve done a good job. Also, given the assumptions that you make, it all adds up.

    Two quibbles – one minor and one major. I thought Rask had a flat out bad year and thought a drop of one position was overly generous. Mind you, I would have kept him in the top ten, so this is just a small complaint.

    I think you underrate save percentage and overrate the new-fangled goalie stats, which I admit I have never actually seen. Save percentage is a fairly high quality stat with far less noise than you may think. It is based on over 1500 shots generally. Certainly, best save percentage will give you the best goaltender with greater certainty than the playoffs give you the best team.

    I agree that all shots are not equal, but I just don’t see that crude measures of differentiation are possible. If you look at the goals that Martin Jones gave up in the SCF, you see a handful of shots that could not have been stopped by anyone. And yes, it is unfair to penalize tenders who see more of those shots. But I don’t believe counting high danger shots and close in shots gets you that number.

    Moreover, I think we need to use the smell test. The new-fangled stats tell us that the Ranger defense was the worst in the NHL and that Hank was an order of magnitude better than anyone else. Who knows, maybe the greatest season ever by a tender. Oh, and Steve Mason is much better than we thought even though his rate stats put him in the lower half of starting tenders.

    When the numbers tell you things that are not believable, you should not believe them. The idea behind the new metrics is legitimate. Some shots are harder to field than others. Unfortunately, subjective judgment (which is sound) will likely vary by observer, while an objective criteria is just hard to quantify.

    Incidentally, personally I don’t particularly disagree with your top choice. Hank is a consistent tender nearly at the top of his game and shows no real signs of aging. OTOH, he was not Vezina finalist worthy last year.

    • Sep 2, 2016 at 12:56 pm

      If you’ve never seen the stats, how can you pass judgment?

      I’d recommend reading up on them, and then revisiting. Otherwise this discussion goes nowhere.

      • Sep 3, 2016 at 2:34 pm

        Dave, some people still think the eye test a better predictor of future play…Hank didn’t pass last year!

        Stats are actually invented to tell any story you would like them to tell.

        A political shell game!

        • Sep 4, 2016 at 10:14 am

          Fair enough. I use my eyes as well, I mean, I do watch 90% of the games for goal breakdowns, make gifs, etc. What I saw was Hank getting hung out to dry.

          Maybe I’m a little biased because I’m a goalie, and I don’t believe the goalie should ever be blamed unless it’s a really bad goal.

    • Sep 2, 2016 at 1:03 pm

      I think those dangerous shots is where you can determine if an NHL goalie is actually better than other NHL goalies. A high percentage of shots NHL ‘tenders see are easy saves, like 19 out of 20 times they should stop them, easy. So the margin is in those few difference making saves. That’s why I feel like SV% is becoming antiquated. Which is hard for me to say because it was just about a decade ago I was laughing at anyone that touted GAA and I would point to SV% as a far better indicator of skill level.

      I guess it just shows us all how quickly quality data and analysis is making it through the sport we love.

    • Sep 2, 2016 at 1:06 pm

      Thanks for the thoughts, Ray. I agree with Dave that I would recommend giving the new metrics their due investigation.

      The other information that I would put out there is that many times, goaltender efficacy and statistical performance may not have that close a correlation. Obviously, hockey is a results oriented business, but there is a very real possibility that, on a micro-level a goaltender can perform well but have poor statistics. Using the example of the PK that Dave had mentioned earlier, those back door, tap in goals, Hank has no chance. However, if he allows two of those in a game, a breakaway goal and then faces 19 other high quality chances, (22 shots total). His save percentage would only be .863% and even his high danger save percentage has been ravaged.

      The question in this example is, did Hank perform poorly? I would argue that he performed well under the circumstances, yet he has been statistically unacceptable in the macro. This is part of what makes statistical evaluation of goaltenders so difficult. You need to look at a combination of factors to reach a more comprehensive conclusion.

      • Sep 2, 2016 at 1:48 pm

        Obviously, you can’t fairly rate a player using the statistics from one game.

        As for the new metrics, I commented above on what I thought were the shortcomings in your ideal stats. Do the new metrics actually take these factors into account or am I judging them fairly?

    • Sep 2, 2016 at 1:58 pm

      5v5 save percentage is decent as an evaluation tool if you believe that the newer metrics still need refinement, which is certainly a fair opinion to have.
      Overall save percentage really isn’t, as the wealth of noise with PK shots tends to reduce how well overall does at predicting future performance.

      Put another way, guys jump between .850 and .900 on the PK all the time, it’s hard to put that on their talent level. But a .930 ES goalie is prone to much less variation in their performance.

  • Sep 2, 2016 at 1:12 pm

    Hank’s #1? Who would’ve thought?

    I realize these are pre-season ratings but Price at #2? Based on 12 Games? Seem like that at least calls calls for an asterisk*

    Just thinking out loud here -But isn’t giving little credence to a Goaltender’s GAA/Save % akin to saying that goals scored by the skaters are not an indication of their value to the team?

    • Sep 2, 2016 at 2:03 pm

      Save percentage is fine. GAA for a goalie is heavily dependant on their work load, for obvious reasons.

      If you have two .920 talent goalies, one faces 20 shots, the other 40 shots, their GAAs are obviously going to vary wildly, even when we know that their talent level is the same.

  • Sep 2, 2016 at 6:56 pm

    Thank you Justin.

    And of course, still the best n the world. Only the Rangers’ D can make the King look ordinary at times. Shame.

  • Sep 3, 2016 at 2:15 pm

    Justo!, Of course you are a Homer! Aren’t we all?

    • Sep 4, 2016 at 10:15 am

      Justin is probably the most level headed of all of us here.

Comments are closed.