Sportsnet.ca http://sportsnet.ca/author/chris-boyle/feed/ Fri, 29 Mar 2024 14:37:50 EDT en-US hourly 1 Getty Images Bishop and Crawford Goaltending no longer key to winning Stanley Cup Wed, 03 Jun 2015 12:25:40 EDT Wed, 03 Jun 2015 15:26:30 EDT Chris Boyle Goaltending has traditionally been the key position for Stanley Cup-winning teams, but as Chris Boyle writes that’s no longer the case, and the Lightning and Blackhawks are a good case study.

]]>
Goaltending is an integral part of playoff success. Every fan has fallen victim to a goaltender getting hot for a couple weeks and dashing potential Stanley Cup dreams. We are only weeks removed from Petr Mrazek almost derailing the Tampa Bay Lightning’s Cup run, and actually Jake Allen derailing another in St. Louis.

STANLEY CUP PLAYOFFS: | Broadcast Schedule
Rogers GameCentre LIVE | Stanley Cup Playoffs Fantasy Hockey
New Sportsnet app: iTunes | Google Play

It is important. The question is whether it important to have a great goaltender or just an adequate one. Even when a goaltender wins the Conn Smythe Trophy, he isn’t necessarily integral to the success of the Cup winner. Jonathan Quick gets a ton of credit for the 2012 Los Angeles Kings Stanley Cup victory, but the Kings didn’t need dominant goaltending to win. I recently referenced a stat by Philip Myrland called the Win Threshold, which essentially measures what save percentage a goaltender would need to produce to finish at .500 or, for playoffs’ sake, make it to overtime. During the 2012 post-season, Jonathan Quick could have produced a nightly save percentage of .894 and the Kings would have played 20 overtime games.

What Win Threshold really indicates is goal support. And what it has shown over the past dozen seasons, unlike in past eras, is that spectacular goaltending is now rarely required to have sustained playoff success.

chart4

If you are leaning on your goaltender heavily, you are likely looking at a ceiling of 8-10 playoff wins at most. Only two goalies have been able to make a Cup Final with little to no goal support during that time frame, J-S Giguere in 2003 (.936) and Roberto Luongo in 2011 (.927) and they both lost game sevens. Of the other 22 goaltenders who have reached the Stanley Cup Final, not one of them required a Win Threshold above .916.

Ben Bishop (.902) and Corey Crawford (.911) fit into the latter category this year.

Neither goaltender has needed to be spectacular to push their teams into the Stanley Cup Final. In fact, both have had performances that place them in the the middle tier of netminders during these playoffs.

chart1

Both have struggled at times, but Bishop (+.009 above expected save percentage) has actually been very good for the majority of the playoffs. Crawford has been bang on the Osgood Line (0.000 differential) and the Blackhawks needed every inch of Scott Darling, the 6-foot-6 playoff saviour (+.007) during the first round versus the Predators.

Outstanding performances by Henrik Lundqvist, Carey Price and Brayden Holtby were not enough to carry their teams to long runs because their teams didn’t provide adequate goal support or carry the play enough.



Download Sportsnet magazine now: iOS | Android | Windows


Bishop has struggled with perception issues because of a tendency to lose concentration on some long shots. This has manifested itself in poor tracking habits with his glove—leading to disastrous results and highlight fodder for the masses to consume.

While Bishop doesn’t possess the technical dominance of Price or the anticipation of Lundqvist, his biggest plus remains his size. The sheer volume of net that he covers doesn’t require him to square up to every shot, but the result is the an awkward way he appears to reach and moves to pucks, and the extended desperation leg stretches he uses when hbeaten laterally.

chart3

Like any goaltender, Bishop is extremely vulnerable to lateral movement and when protected from second chance opportunities he has been significantly above average with a .958 clean save percentage in the playoffs. His success is understandable considering his net coverage. If you allow him to feel comfortable with is backside coverage, he sets aggressive depths which are tough to beat at his size.

This will be something to watch during the Final because the Blackhawks through three rounds are scoring on 7.1 percent of their clean looks, way above the league average of 4.9. Bishop’s size becomes a detriment if you can get him moving by opening him up and creating large holes. If the Lightning can’t limit the lateral passing by the Hawks, they could be in trouble as Chicago has been finishing close to 40 percent of these cross-crease feeds.

Crawford has also struggled with inconsistency, highlighted by his tendency to react down and then out to pucks. Nashville’s Pekka Rinne had similar struggles in round one, as Dan Stewart at InGoal Magazine pointed out nicely. While this does tend to create glove issues, Crawford’s glove-hand weakness has been wildly overblown.

chart2

Chicago has done a good job of insulating Crawford, limiting his exposure to desperation backside recoveries. That’s good, because he has been significantly below average on such shots over the periods I have reviewed. While the Hawks have done well to limit these high quality chances, the Cup Final presents the challenge of taking these opportunities away from the Lightning. They rank No. 1 in creating those chances (7.5 percent of their shots) in the Stanley Cup Playoffs and managed to abuse two of the leagues best in Lundqvist and Price through the first three rounds.

The Hawks and Lightning aren’t here because of their goaltenders, they are explosive offensive teams that only need adequate net support to succeed. Neither team has a clear advantage in goal so the series will likely come down to who can isolate and exploit both goaltenders lateral weaknesses better.

Bishop and Crawford have been prone to explosive meltdowns, so in a small-sample Final it could make the difference. Both have been rather resilient during their teams’ runs, and even though Crawford has the Stanley Cup experience, Bishop has played better.

I think Tampa will be able to exploit Crawford more than Chicago can Bishop and this will ultimately determine the outcome.

]]>
Hockey NHL sn-blog-entry
Bruce Bennett/Getty Tyler Johnson Stanley Cup Playoffs; Tampa Bay Lightning; analytics Lightning’s Triplets Line doing all the right things Wed, 20 May 2015 12:04:42 EDT Wed, 20 May 2015 12:26:52 EDT Chris Boyle Tampa Bay’s Tyler Johnson has caught lightning in a bottle and is leading the Stanley Cup Playoffs in goals and points. But as Chris Boyle writes, he’s defying the analytics to do it. Here’s how.

]]>
After a 72-point regular season, it’s hard to call Tyler Johnson a breakout star, but for fans getting their first extended look at him there’s no doubt he is just that for the 2014-15 Stanley Cup Playoffs.

But his sudden goal outburst doesn’t have him among names like Wayne Gretzky or Sidney Crosby, more like playoff “legends” Ruslan Fedetenko, Fernando Pisani and RJ Umberger. Those are names that scream of a player getting hot at the right time, while also tagged with fluke and unsustainable. Any time players have a shooting percentage in the high 20s and produce at a 60-goal pace red flags are raised for suspect performance.

Now, Johnson isn’t Steven Stamkos, but I’m not sure that this is all luck-based either.

STANLEY CUP PLAYOFFS: | Broadcast Schedule
Rogers GameCentre LIVE | Stanley Cup Playoffs Fantasy Hockey
New Sportsnet app: iTunes | Google Play

The Red Wings gave Johnson a steady dose of Pavel Datsyuk and Niklas Kronwall, and the Habs used P.K. Subban and Tomas Plekanec. I made the assumption last week that the Rangers would match Rick Nash’s line against Johnson’s “Triplets” and use the Marc Staal pairing on defence. But through two games, New York has been using the Derek Stepan line as well as Ryan McDonagh-Dan Girardi against Johnson’s line—the same unit the Rangers used to limit the damage against Crosby and Alex Ovechkin in the first two rounds. That is a ton of respect when you play on a team with Stamkos.



Download Sportsnet magazine now: iOS | Android | Windows


With that type of competition and his current scoring rate, his performance is tough to fathom and definitely seemingly unsustainable. Looking at his shot metrics, it is difficult to understand how he can be plus-five at even strength.

chart1

Johnson is pretty much on par or below the break-even point these playoffs. That makes absolute sense when when looking at his deployment and generally leads back to a conclusion of luck and probability. With an expected-goal total based on location and pre-shot movement we begin to see how he has managed his even-strength totals. Running those numbers against shot-based metrics (which measure each shot as equal) the numbers went from Johnson treading water to Johnson dominating.

chart2

When accounting for important goal variables (pre-shot movement including passes, rebounds and tips) Johnson’s numbers spike to a dominant 60 percent. When plotting the actual shot locations we also see why he and his line are shooting the lights out. When all shots are equal we don’t discount shots from outside the blueline or the exterior low-quality red shots measured versus the high-quality green shots that take up the most quality real estate around the crease.

chart3[1]

Johnson’s line is dominating the highest-scoring area on the ice with high-quality green opportunities—shots with high probability of success thanks to pre-shot passes and blazing speed that force movement by the goaltender. Defensively, Ben Bishop is getting a lots of clean looks and has allowed a couple of long-range low-percentage red goals. An incredible performance from a third-year player being expected to carry a heavy playoff workload.

Johnson also passes the eye test because of his incredible speed and the highlight-reel goals he has scored the past month. This was evident when I reviewed his even-strength shots.

chart4

With the importance of zone entries established, we also see why Johnson is such an effective offensive player. He has controlled the neutral zone with speed and exposing defencemen who don’t challenge zone entries.

Of his 27 shots at even strength, 24 were the result of zone entries, 14 of which Johnson was individually responsible for. He has been attacking the blueline with speed and exposing slow D-men to the outside with blazing net attacks, cutting to the middle to cross-up coverage, creating space for supporting linemates and second-chance opportunities as well as manipulating passing lanes that expose the royal road. It is an effective use of his speed and one that not all speedy players take advantage of.

Is Tyler Johnson a 60 goal scorer? No. He doesn’t shoot the puck enough. He has been full value for his production and fluke or luck isn’t a term I would use to explain his goal-scoring explosion. He has threaded some needles on some incredible shots versus Petr Mrazek in the first round and Henrik Lundqvist in game two, but assuming average finishing abilities on the opportunities he and his linemates have created, I have Johnson at five expected goals instead of eight. If we accept that he has above-average shooting ability, his production isn’t that far off what he should be producing based on his actual play.

The Lightning cannot win the Stanley Cup if Johnson is channeling his inner Umberger. The question is can he continue to produce these types of opportunities against a Rangers team that has been able to negate Crosby and Ovechkin? Johnson won’t continue to shoot 30 percent, but he will continue to produce high-level green opportunities and provide continued cover for Stamkos.

]]>
Hockey NHL TB sn-blog-entry
Ryan Remiorz/CP Subban_1280 How the Canadiens dug an 0-3 series hole Thu, 07 May 2015 15:29:09 EDT Thu, 07 May 2015 15:29:09 EDT Chris Boyle The Montreal Canadiens have consistently won the shot battle, but all those shots haven’t resulted in more goals. Not because they’re unlucky, but because of poor execution.

]]>
Last night’s last-second heartbreaker highlighted the struggles of a Montreal Canadiens team that cannot score. Since Andrew Hammond stopped handing the Habs freebies, the Canadiens have managed nine goals in seven games; their shooting percentage “clipping along” at 3.7 per cent.

The Canadiens were expected to ride the percentages they managed in the regular season and get possession whipped, but the reverse has occurred. Unfortunately, a lot of the analysis has trotted out the old crutch that most fans hate—“bad luck.” Possession has become the proxy for playing well, but possession doesn’t really hold as true in the small-sample-size playoffs compared to the full 82-game season. And even with large samples there are outliers thanks to our lack of proper measurement tools. At this point, we can’t account for these data/knowledge gaps.

The Habs have been carrying possession, but what does that mean? In the first round versus Ottawa, Montreal won the possession battle in four of six games, but they were outplayed overall and once Craig Anderson entered the fray, could not score. The same has held true through three games against Tampa Bay.

The problem is the Canadiens continue to dominate the exterior of the offensive zone, but their commitment to creating offence in straight lines has limited their creativity. Possession doesn’t discriminate against low-quality opportunities, but creating an expected goal total based on location and whether it is high quality (green) or low quality (red) can weed out some of the statistical noise that points to strong performances where one might not exist.

chart2

The Canadiens have consistently won the shot battle, but all those shots haven’t resulted in more goals. Not because they’re unlucky, but because of poor execution.

There is no doubt that shot counts are very influential and help form our perception of events. I looked at every shot from the Canadiens’ playoff games entering Wednesday night, removed all common shots from offence and defence and plotted the remaining. What we see is the difference between what they’ve produced offensively versus what they’ve surrendered defensively.

Of their shots, 147 have an expected save percentage of .975 or better, while they’ve surrendered only 99. If we look at the mapped differential, we can see where these 48 shots are plotted and why the Canadiens continue to struggle to score. They are trading exterior volume for interior quality.

chart4

This was evident when I reviewed the first round and measured clean looks against movement. The Canadiens were the secondd-worst team in creating offensive movement in the round one. The same problem is evident when we measure their power play and it’s stunning lack of movement.

chart1

Through the first round, 62 percent of all power-play goals were the result of pre-shot movement (lateral, tips, rebound) and goaltenders who could set for a shot (red) had a .935 save percentage. When they couldn’t (green) , that marked dropped dramatically to .695. As things stand now, the Canadiens are keeping company mostly with teams who have been eliminated or are about to be.

Their power play lacks the adequate movement and options essential for success. Their most diligent net-front presence is 5-foot-nine. It is tough to take away a goaltender like Ben Bishop’s eyes when he has a nin-inch height advantage. The Canadiens connected on three lateral passes that resulted in shots on goal through their first eight games and their main play is trying to set up P.K. Subban for low-percentage one-timers.

Those factors are why the Canadiens’ possession advantage hasn’t resulted in goals and wins. They have been slightly unlucky, but entering game three, I had them at just over an expected shooting percentage of 6 percent.

With their backs against the wall last night, the Habs managed to put together their most inspired offensive effort of the playoffs since game two vs. the Senators.

chart3

In game three, the possession numbers and overall volume matched the actual performance—Montreal produced offensive movement and did so in higher scoring areas than Tampa. Brendan Gallagher created havoc in front of the net with second-chance opportunities, Jeff Petry was activated and dangerous, and Montreal’s offence created clean looks on just 74 percent of its shots.

But it was too little too late and the Canadiens left their season to chance with zero margin for error. The end result was a back-breaking last minute breakdown and a Tyler Johnson cross crease tap-in.

If the Canadiens consistently out-chanced their opponents and controlled possession, they would be able to survive shooting slumps and poor luck without relying on Carey Price to save the day. But they relied too much on factors beyond their control; playing safe and uninspiring hockey that lacked any semblance of creativity. Their forwards filled lanes like plastic hockey players on rods, putting pucks on net and hoping for bounces and their all-world goaltender to save the day. It isn’t a route to sustainable success.

The Canadiens weren’t unlucky, they were safe. Safe doesn’t win you Stanley Cups.

]]>
Hockey NHL MON sn-blog-entry
Mark Humphrey/AP Ferland1280 Stanley Cup Playoffs; Calgary Flames The Flames will need more luck to beat the Ducks Thu, 30 Apr 2015 16:38:56 EDT Thu, 30 Apr 2015 18:59:03 EDT Chris Boyle The Flames have been the Comeback Kings all season long. But while that can work for a round, as Chris Boyle writes, it’s not likely to last much longer, so Calgary will need to reverse some disturbing trends.

]]>
When Matt Stajan scored what eventually was the winner in Calgary’s game six victory over Vancouver, it served as the perfect microcosm of the Flames’ season. Down big early, out-shot and out-chanced pretty much all game, Cinderella never gave up. Instead they stormed back for yet another third-period comeback—this one maybe their best yet: four goals in 14:43 to win 7-4 and close out the Canucks.

The Flames are a poor possession team—the third-worst in the NHL during the regular season with a 45.1% Corsi For (adjusted for when the score is close). Their Fenwick close number is one of the worst among playoff teams to win at least a round. The Flames eliminated the Canucks despite a significant possession deficit and thanks to a high PDO (101.9). Jonas Hiller outplayed Eddie Lack and Ryan Miller, and secured a date with Hiller’s former team in round two.

STANLEY CUP PLAYOFFS: | Broadcast Schedule
Rogers GameCentre LIVE | Stanley Cup Playoffs Fantasy Hockey
New Sportsnet app: iTunes | Google Play

Although the results were extremely close between the Ducks and Jets, in reality, the Ducks ran roughshod through through games and were fully deserving of their sweep. Anaheim’s offence created havoc for Ondrej Pavelec, allowing him only 81 percent clean looks while providing Frederik Anderson with cocoon like defensive coverage. This lead to the four game sweep and also the highest expected goal differential in round one.

I have previously researched the importance of pre-shot movement and applied it to individual goaltending success, but it can also be applied toward team shooting success.

chart3

Visually charting where the Ducks offence is setting up shop shows why their expected success lines up well with the above chart. Not only are they creating high quality green looks, but they are creating them in the highest valued real estate on the ice. Contrasting that with what they allow defensively shows a team that collapses the passing lanes, clears the front of the net and frees Anderson of cross-crease feeds, tips and second chance opportunities. Looking at these heat maps we see the constant pressure which resulted in the Jets collapsing in every third period of the series.

chart2

At the opposite end of that spectrum are the Flames. While offensively they push the pedal to the medal and create chances off the rush with speed through the neutral zone as Johnny Gaudreau and Sean Monahan create havoc, defensively against the Canucks they were even more generous allowing 81 percent clean looks for Hiller and Kari Ramo.

chart1

It’s a game plan that relies on great special teams and the opposition goaltender gifting goals if you play your opponent heads up at even strength. Hiller outplayed the Lack/Miller combo, but Hiller has to drastically outplay Anderson if the Flames maintain the same porous approach against the Ducks.

This becomes difficult when their goaltenders struggle to contain rebound opportunities like Hiller did in against Vancouver.

chart4

With big-bodied forwards like Corey Perry, Ryan Getzlaf and Ryan Kesler who make a living in the blue paint, the Flames will be hard pressed to stop the Ducks from continuing to dominate the high-scoring areas. You can’t give second-chance opportunities to a team like Anaheim and it will be extremely important for Hiller to absorb pucks and place them in the corners.

Where the Flames can create problems for Anderson is in transition. The Jets only managed nine goals in round one, but five of them were the result of forced lateral movement (3) and deflections (2). If Calgary allows Anderson to settle in and set his depth and angles like he did against Winnipeg on 90 percent of his opportunities, the Flames will not be around for long.


Download Sportsnet magazine now: iOS | Android | Windows


Throw possession out the window. The Flames need Gaudreau or Monahan to play shoot a 50-goal scorer and Hiller to be out for revenge against his former team to have any chance at pulling off the upset. All the signs point to an Anaheim victory and it is the smart play, but the same type of reasoning also had me picking the Flames to finisf first in the Connor McDavid sweepstakes prior to the season.

If I am cheering for anything in this series, it is Brian Burke’s hair, and more games to enjoy it in all it’s glory.

]]>
Hockey NHL ANH CGY sn-blog-entry
Mrazek How the Red Wings beat the percentages in Game 1 Fri, 17 Apr 2015 18:31:48 EDT Fri, 17 Apr 2015 18:31:48 EDT Chris Boyle Red Wings goaltender Petr Mrazek seemingly stole Game 1 from the Lightning, but a closer look at the shot data reveals that Tampa failed to test him enough from the tough areas.

]]>
Playoff hockey always breeds unlikely heroes. Many of these heroes are little known goaltenders who rise from obscurity to lead their teams on unexpected playoff runs.

Last night Petr Mrazek raised his hand and pushed Andrew Hammond aside as this seasons top candidate.

One week ago Jimmy Howard gift wrapped a victory for the Montreal Canadiens and assured the Red Wings of a first-round meeting with the Tampa Bay Lightning. Mike Babcock didn’t offer his vote of confidence and it opened the door for the little known Mrazek to run with the playoff starters job.

I looked at this last year in regards to coaching decisions and how sometimes it is easier to go with the unknown because of the chance of greatness then to roll with what you already know is mediocre. Jimmy Howard hasn’t really proven himself to be among the elite of the league, so Babcock’s decision is totally defensible.

Defensible turned to apparent genius on Thursday. Mrazek seemingly stole a Game 1 win vs the Lightning. But did he really?

We already know how important possession is to playoff success. The way Tampa manhandled Detroit through puck possession in Game 1 generally leads to victory, but one game samples can mess with probability and prediction.

Det_TB_SA
courtesy of war-on-ice.com

Our perception is always influenced by the shot clock. When it reads 46-14 we are always going to assume that an outstanding goaltender performance occurred. A .957 SV% will always carry a ton of weight, but during a viewing of the highlights from Mrazek’s outstanding performance, I noticed a stark contrast from the previous night’s hero, Blackhawks goaltender Scott Darling, who made 3-4 highlight of the night saves.

Possessing the puck to the extent the Lightning did will eventually lead to fortune over the long haul, but firing pucks from distance requires layering bodies in front of the net to alter sightlines, to block a goaltenders vision on pre-shot movement, tip pucks and alter the trajectory and be in position should rebounds drop into the scoring area. When done successfully, the breaks you are looking for will occur.

The problem for the Lightning is they tipped 2 of 46 shots and produced zero rebound shots from a goaltender who was continuously placing secondary opportunities in the slot.

Mrazek_Rebounds

The problem with games like this is that possession has become such a simple way to assess a game, but we don’t really know what to do with the data when we look at this type of performance. So we assume that Tampa had bad luck and Mrazek was outstanding.

The Lightning were a little unlucky, but they also readily accepted the perimeter and didn’t pay the price in front of the net which would have made Mrazek’s life more difficult.

If we look at Mrazek’s ratio of clean looks vs save percentage, we see his workload placed in the high probability of success area. Clean looks on 91 per cent of total pucks is a goaltender playing catch on most opportunities. Clear sightlines. No movement requires perfect shots or relying on goaltender error.

mrazek

Mrazek delivered an above average performance, but this type of environment is not exactly goalie hell. When we breakdown the game using shot location and marking red vs green to highlight high quality vs low quality, we begin to see how this result becomes possible.

Mrazek_Bishop_0000_1

I marked all shot attempts during the game. Pretty easy to see the possession domination by the Lightning when visualized in this manner, but there are a ton of red shots indicating the low value of the real estate the Wings allowed the Lightning to occupy.

Six high quality green opportunities for the Lightning resulted in two goals. The other 64 red resulted in 0. Now here is where fortune comes into the equation. The Red Wings buried the only two green opportunities they had and were rewarded with a gimme by Bishop on a weak shorthanded backhander that squeaked through the five hole.

Mrazek_Bishop_0001_2

Removing the shot attempts still shows a dominant Lightning possession performance, but they still struggled to get to the front of the net. When you remove the shots that have an expected goaltender success rate of .950 or more (I call these save percentage inflators) we see how Detroit managed to beat the odds for one night.

Mrazek_Bishop_0002_3

With all the zone time, the Lightning were able to penetrate the danger zone 13 times, the Red Wings with their limited zone time managed this 9 times. When we move to the highest danger zone (the box directly in front of the crease) the number favours the Red Wings 7 to 5.

None of this is meant to suggest that Mrazek wasn’t fully deserving of the result, he was perfect on his red opportunities, nor is it meant to suggest that possession isn’t the preferred way for sustained success. It is.

The Wings won because Bishop handed them a goal. When they collapsed and took away the passing lanes, the Lightning became content with pounding them from the exterior.

What the Wings pulled off in Game 1 isn’t really sustainable (try telling that to a Washington Capitals fan) and if they don’t adjust, they are likely looking at a quick exit.

The goaltender tends to get the lions share of the credit when the shot clock reads 40+, but Mrazek stole this as much as Pavel Datsyuk did with his 1.000% shooting percentage. The Red Wings won game one riding unsustainable percentages.

Both teams need to adjust. The Lightning need to create some more offensive zone movement and layer bodies in front of the net and the Red Wings need to figure out a way to get Stamkos away from bulldozing Riley Sheahan. If they don’t, Mrazek will have to channel 2010 Halak for the Red Wings to move on.

]]>
Hockey NHL DET TB sn-article
Wilfredo Lee/AP MTL_FLA_2_1280 Why Carey Price deserves the Hart Trophy Wed, 15 Apr 2015 13:49:18 EDT Wed, 15 Apr 2015 15:30:38 EDT Chris Boyle Based both on the stats and his stellar reputation, Montreal Canadiens goalie Carey Price will win the Vezina Trophy. But for carrying a mediocre Habs team, he deserves the Hart as well.

]]>
The Montreal Canadiens’ Stanley Cup hopes rest squarely on the shoulders of their goalie. It has become a repetitive narrative in Montreal because of the legendary Cup runs of Ken Dryden and Patrick Roy as well as the out-of-the-blue Conference Final runs by Steve Penney and Jaroslav Halak.

This season it is 100-percent true. During the CBC’s recent profile of Carey Price, Peter Mansbridge passed on something that Dryden told him. According to Mansbridge, Dryden said “he was a good goalie on a great team. You (Carey Price) are a great goalie on a very good team.”

Dryden is being humble, but he is right in regards to Price. Price is going to win the Vezina Trophy. Although the Vezina is still voted on, it has reverted to its traditional form recently with save percentage replacing goals-against average. Meet the 50 GP mark and lead the league in save percentage and the Vezina is yours.

PRICE5

In eight of the last nine years, the Vezina Trophy winner has met that criteria. The only year that it didn’t was 2007–08, when Martin Brodeur started 77 games and finished just a couple percentage points behind Jean-Sebastien Giguere and Tim Thomas. Add in his superstar reputation and you can see why he got the votes and became the outlier on the list.

Price has definitely earned it—I have data on the last three Vezina winners, and Price’s 2014–15 season graded better than all of them.

PRICE1

So Price will win the Vezina. The real question, though, is does he deserve the MVP?

One of the interesting theories I read years ago in regards to wins was produced by Philip Myrland, a former Hockey Prospectus writer who ran the blog “Brodeur is a Fraud.” He came up with the Win Threshold, which essentially measures what save percentage a goaltender would need to produce to finish at .500. I ran these numbers for Price and the numbers came back at .908.

So if Price and Dustin Tokarski both produced .908 SV%, the Montreal Canadiens would finish with 82 points. Tokarski finished the 2014–15 season with a .910 SV%. His record? 6-6-4. That’s 16 points in 16 games.

Removing those 16 games, that leaves the Canadiens with 66 games where Price impacted the results. Based on their individual-game offensive output and rewarding the Habs with the exact same shootout/overtime success, the Canadiens would have produced 71 points in those 66 starts with average goaltending. For a grand total of 87 points.

PRICE2

The problem is Price didn’t face an average distribution. The Canadiens are not a good defensive team. Last week I looked at Devan Dubnyk and pointed out environment’s role in a goaltender’s success and how clean looks contribute to that success. Price during the 2014–15 season has been able to mitigate the damage that Michel Therrien’s system exposes him to and has carried the Habs to a 10-5-1 record with a .921 SV% when being exposed to 20 percent or more green-level shots.

When I tracked and ran my numbers Price was responsible for 47 goals above replacement level for an expected SV% of .909. When we compare that to the Win Threshold stat created by Philip Myrland of .908, the Canadiens grade out as a mediocre .500 non-playoff team.

In other words, the fanbase should be counting lottery odds and going over new coaching candidates. Instead, the talk in Montreal is of Price joining the immortals that hang in the rafters of the Bell Center by leading the Habs to a possible Stanley Cup.

The problem is that perception and history don’t quite match up when attempting to place Price in the same sentence as Patrick Roy. Not because Price isn’t a great goaltender, but because Roy never had to drag a poor possession team like the 2014–15 Canadiens to a Stanley Cup.

The Canadiens are one of the worst possession teams remaining in the Stanley Cup Playoffs and poor possession metrics generally don’t lead to positive playoff results. Even through the use of two-period shot metrics (used because there is no historical tracking of blocked and missed shots or html data sheets to separate score effects) we see that outside of the 1980s that poor possession is unlikely to win you a Stanley Cup.

Making the assumption that the Canadiens are likely to lose the possession battles in most of the series they play in, I wanted to contrast how successful Roy was during his playoff career when faced with an opponent outplaying his team. Using the two-period shot metrics available (note: not every year had full boxscores, so I only referenced the available information) at the Hockey Summary Project I checked Roy’s resume against his legend.

The biggest of the Roy legends—how he placed a mediocre Canadiens team on his back and dragging them to their last Stanley Cup in 1993—ignores that they were an above-.500 possession team at .513 and that during their playoff run they held the possession edge in three of the four series they played. The same held true during the 1986 Cup run.

PRICE3

None of this implies that Roy wasn’t great or deserving of the Conn Smythe in either of those seasons, but he didn’t produce those Cups alone. He managed to dominate, but it was behind a strong team with structure. Interestingly enough, the one year where Roy didn’t win the Conn Smythe Trophy during one of his Cup runs was the season where he probably pulled off his greatest performance.

PRICE4

In 1996 the Avalanche knocked off the Detroit Red Wings. While the Avs were a strong possession team themselves at .541, they got bludgeoned by the Wings who were a .579 possession team coming off a 131-point season (pre-loser point) and had a +144 goal differential.

While Roy produced playoff victories while being outshot, when I separated his success into positive and negative possession there was a stark contrast. Roy played in 12 playoff series where his team lost the possession (shot differential) battle. His record in those encounters was 7-5. When his team held the possession edge, it jumped to 18-3.

Expecting Price to pull off a feat that Roy never really did in Montreal isn’t fair. It’s also part of the hyperbole that makes being a goaltender in Montreal so difficult.

Is it possible for Carey Price to drag the Canadiens to a Stanley Cup? Sure. Before the 1993 playoffs began if we discussed the probability of a team winning 10 straight overtime games we would have concluded it as virtually impossible. Before last season, no team had every made the Stanley Cup Final after playing three straight game sevens. Not only did the Kings make the Final, they won the Stanley Cup. So while the odds are against a poor-possession team winning the Cup, anything is possible.

If Price can continue his regular-season dominance through the playoffs, then he will have earned the legend that comes with it. All the odds point against it, but that’s what makes the Stanley Cup playoffs so much fun.

]]>
Hockey NHL DET MON sn-blog-entry
Billy Hurst/AP dubnyk Examining Dubnyk’s Wild comeback season Fri, 10 Apr 2015 14:34:24 EDT Fri, 10 Apr 2015 15:28:20 EDT Chris Boyle A lot of hockey fans want to thank Sean Burke for Devan Dubnyk’s turnaround. But how much of it is thanks to a supporting defensive structure and getting clean looks at shots?

]]>
The first nine months of 2014 were difficult for Devan Dubnyk. While hardship followed his departure from Edmonton, he didn’t realize at the time that he had actually been rescued from a firing squad.

Last season was a disaster for Dubnyk and one that I highlighted in November while looking at the environment that lead to Ben Scrivens’s struggles this season.

Scrivens experienced Dubnyk’s journey in reverse as he went from the highs of a strong team structure in Los Angeles to the chaos of a defence lead by Justin Schultz, but there are parallels. It is easy to understand how things can snowball when forced into recovering from continuous broken plays and the difficulty of maintaining the discipline not to cheat. Dubnyk looked lost at the tail end of his Oilers career and seemed to have lost all confidence as he bounced from Nashville to Montreal to Hamilton.

And then goalie voodoo.

Theories began to point to Sean Burke being a goalie whisperer and turning Dubnyk’s game around upon arriving in Minnesota by teaching him the inside-out approach to goaltending preferred by Henrik Lundqvist and Mike Smith. I don’t doubt that Burke helped in rebuilding Dubnyk’s confidence, but if the inside-out approach is the cure-all for large goaltenders, then Burke needs to whisper some more into Smith’s ear because he has been struggling badly since his one great season in 2011–12.

The second theory—and one more structured in logic and goaltender education—was Dubnyk being taught a new technique called head trajectory by former New York Rangers goaltender and current MSG personality Stephen Valiquette. It’s a new way of tracking the puck and was developed by Lyle Mast, the founder of Optimum Reaction Sports.


Sportsnet Magazine Stanley Cup Playoffs
Edition:
The six reasons why Carey Price can take the Montreal Canadiens all the way. Download it right now on your iOS or Android device, free to Sportsnet ONE subscribers.


Technical greatness should be the goal of every single NHL goaltender who straps on pads, but I have witnessed goaltenders who are technical messes like Tim Thomas produce a .938 SV% and Carey Price, considered the greatest technical goaltender in the NHL, drop a .905.

So I look to environment.

While reviewing Dubnyk’s last two seasons, I noticed the same kind of extremes as Scrivens. Interestingly enough, Dubnyk’s numbers in Edmonton were significantly below the expected .905 and his results in Minnesota were way above the expected .922, but when the two season samples (including his time in Arizona) were merged, the large sample placed him at slightly above average.

We can see visually why Dubnyk struggled in Edmonton. I have slightly altered my heat maps to separate all pre-shot movement against clean shots, creating some consistency with the terms that Stephen Valiquette has been using through his work at MSG. The green area represents higher-quality opportunities and the red area lower-quality opportunities.

Dubnyk_Workload

In Edmonton, Dubnyk was bombarded with high-quality slot opportunities. Forced to set and judge depth while on the move is a recipe for disaster and Dubnyk—like Scrivens and Viktor Fasth—wasn’t equipped to handle this type of volatile environment. Just over 80 per cent of Dubnyk’s looks in Edmonton were clean. Contrast that to Minnesota where his clean looks went up to 88 per cent and Dubnyk’s size became an asset again. This accounts for the .017-point differential in expected SV% differential.

When I started this project, one of the things I did was track individual games and the effect clean looks had on save percentage. The more clean looks, the more success.

Dubnyk_SV

Dubnyk is no different. If he gets defensive support and is allowed clear sight and the ability to set his depth and angle, he has more success. When the defensive structure in front of him clears shooting lanes and clogs passing lanes, Dubnyk is a .937 goaltender—when they fail to do so he becomes an .885 goaltender.

We make assumptions that goaltenders control their destiny, and to an extent they do. The position is a random set of math problems being fired at them at 100 m.p.h. and their job is to judge distance, angle and velocity, and attempt to read and recognize the problem scenario before it happens. It is why the mental aspect of the game is so important. If you have all the technical greatness in the world and your recognition is poor, you will position yourself for a scenario that isn’t going to occur. The elite goaltenders in the NHL are able to do both and get maximum coverage for most scenarios even when things breakdown.

The problem: Some goaltenders get more fifth-grade-level math problems and some are tasked with solving the Kobayashi Maru. We can see this with Dubnyk. When we separate the games where he failed and where he excelled we see that success in these scenarios isn’t random.

Dubnyk_GoodvsBad

The “good start” and “bad start” labels imply that the goaltender is fully in control of the result. Over the last two seasons I separated the shots when Dubnyk was considered to have been successful and starts where he would have been attached blame for failure. The chart on the left is below-average results—the chart on the right above average. Just like the Edmonton/Minnesota split, the more pre-shot movement he faced, the less success he enjoyed. The more clean looks from the perimeter he faced, the higher his save percentage.

Through two full seasons of advanced goaltender data, Dubnyk grades out around average, right in the NHL middle-class. This matches up closely with his career save percentage of .914.

The important question that can be answered as I gather more data is whether the Oilers were a .905 team while he was producing save percentages of .914 to .920. If Dubnyk was actually a +.010 above-expectation goaltender, then he was merely a small-sample failure on a team lacking defensive structure and is everything that management should be looking for in exploiting a market inefficiency.

The Minnesota Wild admitted that the Dubnyk acquisition was born of desperation, but identifying players like this is the driving force behind analytics. The first team to separate the defensive noise from actual goaltender performance will have a window where they can pick and choose undervalued goaltenders on the NHL market as well as raid major-junior at minimal cost.

]]>
Hockey NHL EDM MIN sn-blog-entry
Jeff Haynes/AP sharp_patrick Sharp’s slump goes way beyond bad luck Mon, 16 Mar 2015 12:48:59 EDT Mon, 16 Mar 2015 17:15:32 EDT Chris Boyle Patrick Sharp’s shooting percentage has plummeted this season, but the Chicago winger’s recent slump can be blamed on more than just bad bounces.

]]>
With all the great work being done in the NHL analytics world, it’s frustrating to continually see intelligent hardworking analysis tripped up by faulty NHL data. Jennifer Lute Costella (JenLC) is one of my go-to bloggers when it comes to the Chicago Blackhawks, and she recently put together an expansive breakdown of Patrick Sharp’s 2015 scoring struggles.

Diving into the rabbit hole that is war-on-ice.com, she argued that this slump was likely based on Sharp stringing together some poor shooting luck.

The problem, as always, is the reliability of the data we are analyzing.

SHARP

A review of Sharp’s stats made it clear that his struggles came at even strength, and it’s easy to see where and why the assumption of bad luck entered the picture. Sharp’s goals are down and his PDO points to extremely poor luck based on his poor shooting percentage. His zone usage, scoring chances and possession numbers all match up with his previous season, where Sharp shot an above average 9.5 percent.

This is where things got interesting, as Darryl Belfry, an NHL skills development coach who has worked with Patrick Kane, entered the conversation.

“So what’s wrong now,” Belfry wrote, “well injuries derailed his momentum and mobility. As he worked his way back into he lineup he couldn’t find his timing. In reaction to lack of production Coach Q drops him in the lineup. Once he plays a few shifts outside a line that doesn’t have 88 or 81 the possession timing is thrown further out of whack. Now instead of sliding into seams, he is ‘standing around’ with longer and longer stretches of time where no passes are coming his way in those seams. In reaction to ‘standing around’ and the frustration of not getting the puck in those seam windows, he compensates by starting to ‘chase the puck’ for contact more to force more turnovers. Now instead of extending possession sequences he is in never-ending cycle of going from one board battle to the next.

As I reviewed and tracked Patrick Sharp’s last two seasons some of Belfry’s observations jumped out at me. Once Sharp was separated from Hossa and Toews, most of his even-strength minutes came alongside Andrew Shaw and Brad Richards. While I expected more pre-shot movement was the answer, neither season saw Sharp receiving one-timers or cashing in rebound opportunities. What I observed was Sharp not getting the puck in stride in transition. It affected his ability to fill seams and instead he became the main puck handler attacking the blueline.

In 2013-14, Sharp was able to free himself for 12 breakaways, the 2014-15 season has seen that number drop to two. Without Hossa and Toews feeding him in stride and opening up space for him, Sharp has maintained a high shot count, but the majority of the shots have come after crossing the blueline, pushing wide and firing pucks from all angles.

This is visually evident in his expected shooting percentages.

Sharp_Exp_740

Remove the white noise of luck, and it becomes clear that Sharp just isn’t generating the same type of opportunities he did in 2013-14. Last season, he converted even-strength opportunities at a slightly above average rate. In 2014-15, he hasn’t met expectations.

Here, we run into collection issues again, as the tracking data provided by NHL.com is spectacularly inaccurate and lacks pre-shot movement.

Here is a side-by-side of Patrick Sharp shot charts, one assembled by visually tracking the co-ordinates through the use of NHL GameCenter and the other based on current logged locations courtesy of War-On-Ice.

P_Sharp_Both_740_2

The NHL.com data inaccurately shows Patrick Sharp penetrating the home plate area, while visual tracking and research paint an entirely different picture, showing a player who is being contained on the perimeter because he is forced to shoot off the carry.

Over and over Sharp gains the zone. If the D provides him the spacing to shoot, he fires from above the hash marks. If he’s pressured, he swings wide and fires from bad angles.

The difference in his shooting percentage becomes glaring when we compare and contrast it to the previous season, where he was able to maximize his asset base with Hossa and Toews.

P_Sharp_14v15_740

When we remove the noise created by inaccurate data, we begin to see a reason why his shooting percentage has crashed—beyond luck. In 2013-14, Sharp was consistently gaining the home plate area for high-quality opportunities. This year, those opportunities have virtually disappeared, which matches up nicely with the eye test provided by Belfry.

“What’s wrong with Patrick Sharp is his current game habits are in conflict with his asset base,” Belfry writes. “He is not a player who can dictate conditions on the ice, he is a partner on a line not a centrepiece. Must play with the right type of player—elite possession player—whom he can read their possession habits and dial in his timing with so he can do what he does best … Patrick Sharp is an unreal talent, however, like many players who play in the NHL, he must find a way to adapt his game to changing personnel, role and ice time and find a way to continue to stay true to his asset base.”

Sometimes, in our rush to provide counter-intuitive observations and arguments, we ignore the answer right in front of us. Sometimes a slump isn’t the result of poor play, but rather of the environment a player has been placed in. I referred to this in regards to Ben Scrivens early in the year, and Alexander Ovechkin last week.

Patrick Sharp remains a great player, but he needs to be placed in a situation that allows him to succeed. The Blackhawks simply haven’t provided that this season.

]]>
Hockey NHL sn-blog-entry
Paul Vernon/AP WSH_CLB_1280 Ovechkin not to blame for poor plus-minus Mon, 09 Mar 2015 14:30:24 EDT Mon, 09 Mar 2015 14:30:24 EDT Chris Boyle Alex Ovechkin has a bad rep when it comes to defence—and last year’s epic -35 didn’t help. But poor coaching and goaltending are more to blame for the perception than he is.

]]>
Alex Ovechkin’s even-strength play has always been a target of criticism. Whether it is his perceived defensive indifference (this gif obviously doesn’t help his cause) or last season’s infamous minus-35, Ovechkin has always struggled with the “poor defender” label.

When the Capitals fail to win in the post-season, the conversation generally moves to Ovechkin. And since he’s been essentially the same offensive player in the playoffs as in the regular season, the easiest attack points become his character and again his lack of defensive commitment.

As advanced analytics attempt to combat this surface analysis, the same old trope is continually beaten to death: Hockey is a complex game and there isn’t a single stat—like baseball’s WAR—to simplify and ascertain a player’s effectiveness. But that’s precisely why a simplistic stat like plus-minus became a go-to for the lazy, even though it lacks context and can be influenced by uncontrollable factors like coaching tactics.

When Ovechkin was in the middle of his horrific 2013–14 season he was hammered for his minus-35, but the criticism totally ignored the environment in which it was produced. At no point did anybody place any of the blame on Adam Oates. Never mind that prior to Dale Hunter, Ovechkin had been plus-88 for his career. When Hunter was hired, the story became how his tough approach was forcing Ovechkin to block shots and was transforming him into a more defensively responsible player, even though none of the stats confirmed this.

Hunter did start Ovechkin in the offensive zone less, which impacted his numbers, but while doing so he managed to nullify one of the greatest offensive players in history. Oates replaced Hunter and reinvigorated Ovechkin’s power-play production, but even with favourable zone matchups his 5v5 play continued to collapse.

OVIE2

We’ve recently seen in Toronto how a new coach and a change in tactics can affect individual possession results. And with Barry Trotz replacing Oates, Ovechkin has returned to being an even-strength force.

Plus/minus also has a significant blind spot because it can be influenced by dominant or poor goaltending as well as percentage-driven results. I reviewed Ovechkin’s even-strength play (5v5 and 4v4) and attempted to figure out what his results would be if he received league-average shooting and goaltending.

(Note: When I see massive year-over-year contrasts in results, I check them vs. my research to see if I can clarify the conditions and the environment in which they were recorded. I don’t automatically put the blame on the goaltender—I want to check the context to understand the reason for the huge discrepancy.)

One of the most interesting things I noticed when dissecting Ovechkin’s performance under Oates was how percentage-driven both years were. The shot rates against were almost identical.

OVIE1

During Oates’s first season, the Capitals’ goaltending and shooting aligned with their expected result. This resulted in the expected +/- matching up with the actual result of +3.

The 2013–14 season was an extreme contrast. Defensively the Caps’ shot rate against got worse, but their expected save percentage remained consistent. The issue was the collapse of the Capitals’ goaltending. The backup-level netminding resulted in eight more goals against than expected while Ovechkin was on the ice. Offensively Ovechkin and his linemates produced lower-quality opportunities and cashed in fewer than the season before. This led to another 10 offensive goals that went unconverted.

Even removing these percentage-driven results, the Capitals under Oates still had a negative differential with Ovechkin on the ice at even strength. Instead of a 12-goal drop-off between seasons and a -9, the result was a disastrous 30-goal differential and a -27.

With the hiring of Trotz, Ovechkin’s offensive percentages returned to normal.

OVIE3

Under Trotz we can see the improvement. Ovechkin is relying less on clean looks to beat goaltenders and his line is employing more pre-shot movement

Defensively his line is suppressing shots and when those shots are registered, they are lower quality. Almost 90 percent of the shots are of the clean variety and they have cut the rebound opportunities by more than half. For the first time in three seasons his expected on-ice save percentage moved past the league average of .922 to .923. Unfortunately the Capitals’ goaltending again has underperformed.

Although backed by sub-par goaltending again, their improved play has avoided the disastrous negative goal differential of the previous two seasons when they became overly reliant on percentage-driven success.

Trotz walked into a great situation where everybody believed that Ovechkin was broken and needed to be fixed. In the end, that was far from true. All Ovechkin needed was an environment in which he could succeed.

]]>
Baseball Hockey NHL sn-blog-entry
Trevor Hagan/CP ovechkin Winnipeg Jets; Pavelec; analytics; shot quality; Hutchinson Jets’ Pavelec relies on the big save too much Mon, 02 Feb 2015 13:35:33 EST Mon, 02 Feb 2015 13:42:26 EST Chris Boyle Many would look to Ondrej Pavelec’s highlight-reel saves this season as an indication that he’s working his into the category of elite goaltender. But Chris Boyle writes that’s hardly the case.

]]>
The challenge in analyzing Ondrej Pavelec’s career lies in the influence of Sportsnet Central and the like over the general fan base. When exposed to the imagery that populates great-save reels, we are generally treated to goalies who have placed themselves in dangerous positions that force them to rely on desperation to make miraculous saves. Every goaltender in the NHL makes these saves, but the question lies in their importance.

The issue is about opportunity. Netminders who trail plays and generally find themselves out of position more often increase their opportunity for big saves. More chances for such saves mean more opportunities to populate highlight reels, which work as currency for fans’ opinions on goaltenders they don’t regularly follow. It can be just as influential as save percentage for the fan who tracks performance through data.


ATTENTION Rogers and Shaw customers.
Already enjoying Sportsnet ONE? Now get access to digital editions of Sportsnet magazine at no extra cost.


That is why Pavelec is so polarizing. When a fan sees a performance like the one Pavelec delivered against the Dallas Stars on January 15 it can be taken as evidence of his greatness. A typical average expected save percentage of .913 would have predicted four goals against Pavelec. When I ran it through my location/movement data it rose to five goals against or an expected SV% of .894. During his outstanding performance against the Stars, 12 of the 47 shots he faced were high-end with pre-shot movement. He also faced a breakaway and finished the night with a .979 SV%.

It’s performances like that one that show how single-game samples wreak havoc on the eye-test analysis. When contrasted with other mediocre goaltenders, those who support Pavelec’s overpriced contract often fall back on his save percentage being the result of a poor team in front of him. It is an argument that has merit in certain circumstances and one I argued in favour for in regards to Ben Scrivens in Edmonton.

And with stats we can propose the team argument by looking at how his peers have performed as Winnipeg Jets/Atlanta Thrashers since 2008.

chart3

From that data, we can infer that the team has been bad for years, hence continually poor save percentages for its goalies. Unfortunately, outside of Kari Lehtonen, the goalies being referenced are mostly unproven NHLers or career backups—guys like Johan Hedberg, Chris Mason, Peter Mannino and Al Montoya. So in reality, Pavelec—along with his $3.9-milllion cap hit—is actually being compared to borderline NHLers, not bonified No. 1 starters.

The question then is whether Winnipeg/Atlanta goalies have all been significantly below league average or if they’ve been dragged down by a poor defensive structure. I ran 2,200 shots through the shot quality machine and got the following results:

chart1

The Jets have been slightly below average defensively with an expected SV% of .911 with Pavelec in net—below average, just what haunts Pavelec. He is exactly league average on clean opportunities, but his struggles are exposed in situations where he is overactive or is late tracking the puck. His save percentage plummets when fighting through screens. While reviewing him, I have noticed a lack of willingness to maintain visual attachment to the puck. When he loses sight, he has a tendency to guess and drop into the butterfly. That guessing leads to overactive responses on simple plays, meaning he’s the cause of a lot of his own issues. And when he is able to clean up his own mess, it is generally in an inefficient manner, which results in a spectacular save.

Here’s a series that shows exactly what I mean:

Pavs_1

Above we see Pavelec losing visual attachment to the puck and not fighting for his sight lines. Here he is tracking Ryan Getzlaf across the royal road. As Getzlaf recognizes Ryan Kesler open for a backdoor feed, he begins to reverse the flow. At this point Pavelec is in good position.

Pavs_2

But once Getzlaf has reversed the flow and the puck is on route to Kesler for a one-timer, Pavelec has lost sight of the puck. He guesses a shot is on the way and drops into the butterfly.

Pavs_3

As Kesler begins to release the one-timer, Pavelec has recognized what Getzlaf has done and is now searching for the puck. At this point Pavelec is in panic/desperation mode and must rely on his reflexes, which is generally when the spectacular save emerges.

Pavs_4

Except there is no miraculous recovery. As the shot crosses the goal line, Pavelec has finally read the play. This scenario can easily be dismissed as a lateral feed where the goalie has zero chance. It is a situation that involves one of the highest probable scenarios for offensive success: traffic in front, a man advantage and a lateral back-door feed across the royal road. But elite-level goaltenders fight, read and assess these situations earlier than Pavelec. When you repeat these behaviours over and over, they result in higher levels of success in larger samples.

Pavelec is trailing the play and goaltenders who do that with regularity need to be in controlled and protective environments. When they are, their flaws can be hidden because they aren’t forced to assess difficult scenarios as often.

Fortunately for the Jets, they have a goaltender who plays a more controlled style and—even as a rookie—rarely trails the play in this manner. Michael Hutchinson has been impressive during his small sample success. When contrasting his play with Pavelec’s this season, it’s clear which one should be getting the bulk of the starts.

chart2

The question becomes can Hutchinson maintain this level of play. It is unlikely that he is a .930+ goaltender, but the competition is replacement level. Hutchinson is significantly above average in every category I track except for rebound shots, where he is -.013 below the league average. The number that will likely drop is his save percentage on clean shots. Hutchinson has been successful on over 96 percent of the shots he can see and set for. This makes up 85 percent of all shots he faces.

Even if we went to extremes and dropped him to the league average on clean shots, he would still rate comfortably above Pavelec’s below average career.

I am not a fan of scanning the save percentage stats and pointing at goaltending as a resason for team struggles without context through deeper research. In this case, it is correct to assume that if the Jets want to make a real run for the playoffs, they would be smart to flip their goaltending usage rates and use Hutchinson as their guy over Pavelec.

]]>
Hockey NHL WPG sn-blog-entry
Nick Wass/AP nash_rick640 Rick Nash; anlaytics; New York Rangers; NHL All-Star Game; Where Rick Nash’s Rangers goals are coming from Thu, 22 Jan 2015 16:31:18 EST Thu, 22 Jan 2015 16:54:49 EST Chris Boyle Rick Nash is on pace to have a career season scoring goals. But the question is why? Chris Boyle has dug in and tells you what is going on—and whatever you do, don’t chalk it up to good luck.

]]>
One of the main reasons I began the shot quality project was to fill in major gaps of information missing from NHL data reports. Through my research I discovered inconsistencies when it came to shot locations coupled with significant information missing on pre-shot movement. In short, a lot of conclusions were being made without the correct information.

The analytics community is filled with intelligent individuals who come up with logical conclusions based on the information at their disposal. The problem with the gaps in information is that we begin to lean on luck because we just can’t logically explain the reasons without making assumptions based on intuition or research based on this same limited data.

War on ice as emerged and they are attempting to fill in these gaps. Their charts and scoring chance accumulation have provided great new resources for analysis. The site is a great resource and would be an even greater accumulation of data if NHL.com did not have accuracy issues.

Trying to assess how Rick Nash is producing at a 52-goal pace is futile using numbers from war-on-ice. His shot rate hasn’t changed and his scoring chances for are in line with the previous two seasons in which he scored at 33 and 39 goal pace. His breakaway opportunities are stable—about one every 10 games—so that’s not it. Simple solution: Luck. His shooting percentage is at a rate he hasn’t seen since his 21-year-old season.

nash3

Looking at Nash’s numbers since he joined the Rangers, he isn’t producing at a rate that would suggest a 50-goal season. His shot rate is actually slightly down from 2013-14. You could conclude luck, but the term itself means different things to different people. Some take it to mean lack of skill and entirely due to fortune. It can also mean pucks finding players in good positions on the ice, lucky bounces, or even bad goaltending.

I prefer to search for logical and more concise explanations. Steve Valiquette of MSG Sports has also been doing shot quality analysis and has done a fantastic job of visualizing the work. He has a slightly different evaluation than mine and has added the concept of the royal road to the discussion. Earlier in the season he reviewed Nash’s success and concluded it was based on pre-shot movement.

Similar to my philosophy that success is earned by making goaltenders move laterally, I decided to review Nash’s Rangers career. Highlighting the difference between a shot from the exterior versus a pass across the royal road we see a significant increase in goal probability when Martin St. Louis unleashes a goal-mouth feed instead of a harmless exterior shot.

nash1

With the pre-shot movement, St. Louis takes goaltender James Reimer from a situation where he would be expected to succeed 98 out of 100 times to one where 60 out of 100 times is an average.

Nash’s first two seasons in New York saw him take clean shots 90 percent of the time. His conversion rate on those opportunities was 10 percent. With the Rangers, his finishing rate on shots with pre-shot movement is 32.5 percent. But in the past two seasons, Nash only managed to take 44 such shots out of the 429 he registered. During this time, Nash scored 37 of his 47 goals through clean shots. There’s no doubt Nash is a sniper—you don’t need advanced analytics to understand that—he is a career 12.7-percent shooter and doubles the average success rate on clean shots. So what’s different this season?

nash2

Of the 156 shots Nash has taken in 2014-15, 36 have included pre-shot movement (lateral passes, deflections, rebounds). Twenty-three percent of his shots are of a variety where he has a 32.5-percent finishing rate. The result, 16 of his 28 goals are thanks to these higher-end opportunities.

The question is sustainability, individually and team wise. It is unlikely that he continues the 44-percent rate he is currently scoring at on these opportunities. His three-season conversion rate is 32.5 percent, and his previous two-season rate was 23 percent. Also, will the opportunities continue at the same high rate? St. Louis has been responsible for a third of the passes that preceded Nash’s shots, but Nash spends the majority of his time with Derrick Brassard and Mats Zuccarello. Nash is also more active in front of the net, where he has doubled his tip opportunities and has a higher frequency of second-chance rebound shots. Will he keep that up?

A slight decrease in opportunities or finishing rate and Nash will fall short of his first 50-goal campaign.

The point of this study is to add a new layer of context to previous observation. While we may be forced to guess the reasons for his career bump because of the limited NHL.com data, more information allows for a rational explanation for his production increase. Nash is scoring more because of the quality of his shots.

]]>
Hockey NHL NYR sn-blog-entry
Jason Franson/CP ben_scrivens In the NHL, save percentage is a team stat Tue, 25 Nov 2014 10:14:09 EST Tue, 25 Nov 2014 13:32:55 EST Chris Boyle The Edmonton Oilers fired goaltending coach Frederic Chabot Monday. But Chris Boyle points out that goalie success in the NHL is really a team stat, not an individual one.

]]>
Once again Monday, a goaltending coach was scapegoated for a team’s perceived poor netminding. But does anyone really believe that Frederic Chabot was the reason the Edmonton Oilers have the NHL’s worst team save percentage? For that matter, are Ben Scrivens and Viktor Fasth even to blame?

The goaltending position continues to be misunderstood. Goalies have one job: Remain as square to the puck as possible. They can control their reads, decision making and movements to the puck to beat the pass and make the save—or solve the equation—but have zero control over the difficulty of the equations thrown their way. That’s why I am never surprised by radically fluctuating save percentages; I believe that environment influences numbers.

One issue I take with analytics is the notion that large samples solve everything. That doesn’t makes much sense. It assumes that over the past four or five years, a netminder on the Bruins (with Zdeno Chara) has seen the same scenarios as a goaltender on the Islanders. Or that if Chris Kunitz played with Tyler Bozak instead of Sidney Crosby, he would produce at the same rate because the larger sample evens out. Nobody believes that, yet without any reasonable explanation, it is a common assumption in analytics that Tuukka Rask is 10-to-12 goals better than Carey Price per 1,000 shots.

That is the level of analysis we are currently seeing. I touched on this peripherally last season when looking at Pat Burns and other defensive oriented coaches. During his Montreal/Toronto years, Burns consistently harped on controlling the neutral zone. Turn the puck over at a blueline and you would find yourself in big trouble.

The New Jersey Devils under Jacques Lemaire were also consistently disciplined in that highly influential patch of ice. Strong save percentages followed those two coaches around. With the current focus on zone exits and entries, and their link to better opportunities when teams cannot force dump-ins, it makes sense that the neutral zone was so important to those coaches.

The idea that zone entries influence offence/defence/possession is backed up by the data. It doesn’t matter if the x,y co-ordinates are correct or not, it’s a fact that more shots are created on controlled entries versus dump-ins. Zone entries are also linked to shot quality. If a team can control the neutral zone, excel at zone exits to negate efficient forechecks and limit controlled entries, it should be able to stifle offence.

Can a team do this consistently in a large enough sample to limit its opponents’ offences and in turn inflate its own goaltenders’ save percentages? Yes. Case in point: the Boston Bruins.

bruins-habs

I tracked three full Bruins seasons—every shot faced from the 2011-12 through ’13-14 by Tim Thomas, Chad Johnson, Anton Khudobin, Marty Turco and Rask. I also included Rask’s 2010-11 campaign. Over the 6,700-shot sample, the expected save percentage for an average Bruins goaltender was .923. The only sample I had large enough to compare was the Montreal Canadiens—a 4,800-shot sample of Carey Price, Peter Budaj and Dustin Tokarski. Their expected save percentage over the same period was .908.

Using the percentages and extrapolaying the shot totals to 6,700, the Bruins goaltenders over this sample would include almost 400 more clean shots—which have an expected save percentage of .949 rather than the .651 on shots preceded by a pass. Since 2008-09 the Bruins have maintained a .926 SV%. Plug in a goaltender and a .920-plus SV% shows up. Why? Because the baseline for average on the Bruins is not the league average of .914 it is actually .923.

When we look at the Bruins having an increased baseline average then we have to re-categorize their goaltenders. Instead of six goaltenders significantly above average, we have Rask and Thomas above average and Johnson, Khudobin and Svedberg average or slightly below. That’s a more logical theory than every goaltender in Boston is really good. Price who is widely regarded as the blueprint for goaltenders, struggles to crack the league-average save percentage at .917, but if his baseline is .908 it passes the eye test for the most technically efficient goaltender I have ever seen.

Instead we are bombarded by a lack of understanding of a complex position simplified to one statistic—save percentage—that entirely ignores individual environment. Ondrej Pavelec gets off to a strong start and the talk turns to his adjustments and enhanced play instead of looking at how his environment has changed.

jets

Simply put, 27-year-old goaltenders who over-pursue pucks and have struggled to keep up with the pace of play for 290 NHL games don’t adjust to the speed of the game over an off-season. However, their environments can be altered. Early in the 2014-15 season, more than 90 percent of the shots Pavelec faced were clean. My research indicates that he has an expected success rate of 95 percent on such opportunities. Previous data had his clean shots faced at 84 percent.

The changing environment theory is best exemplified by goaltenders who move from team to team. Ben Scrivens spent three seasons trying to establish himself as an NHL starter. Over a 6,000-shot sample in college, the AHL and ECHL, Scrivens registered a .928 SV%. Pre-Edmonton, he had a respectable .917 NHL SV% over close to 1,400 shots. During his small-sample stint in L.A., Scrivens led the league in save percentage. But since his trade to Edmonton he has cratered, posting a disastrous .905 SV% (.887 this season). And it’s entirely environment based.

scrivens

A the chart above shows, Scrivens is a respectable +.001 above expected save percentage for his career. But it also shows what’s actually been expected on his different teams—922, .919, .909—and helps illustrate the massive difference in workload he has now compared to his first two NHL stops. Scrivens’s career .957 SV% on clean shots coupled with the Kings’ ability to limit high quality pre-shot movement opportunities was a perfect match of styles. Even in Edmonton, his save percentage on clean shots is +.011 above expected, but his exposure to high-quality opportunities in the form of cross crease passes has buried him. In Edmonton, Scrivens is facing 30 to 40 more difficult shots per 1,000 than he was in Los Angeles, or even Toronto.

Ultimately the Oilers’ struggles cost goalie coach Frederic Chabot his job. But the real reason Scrivens’s save percentage is below .900 is porous defensive zone coverage. When running the data for just 2014-15, Scrivens has an expected success rate of .903. Only 82 percent of the shots he has faced are ones in which he could set his depth and angle. The Bruins register 88 percent.

Scrivens spent close to a decade being above average before arriving in Edmonton, a team whose goaltenders have registered one season above league average since 2009-10. Eleven different netminders and the only ones above the Osgood line (league average) played three or fewer games. Only Devan Dubnyk (.910 to .909), Martin Gerber and Richard Bachman (who both faced fewer than 100 shots as Oilers) were better in Edmonton than elsewhere.

This is environment at work not goaltending ability. How is it possible to have any confidence when you can’t stop pucks thanks to the impossible equations being thrown your way? Desperation can cause goalies to lose confidence in their process and start chasing the game. Instead of hanging in for shots, they cheat because they’ve lost trust in ther teammates. The same thing can benefit goalies in a controlled environment. Jonathan Quick and Tim Thomas use more aggressive depths because they’re not concerned with backside pressure—they trust their teammates. It snowballs in both directions.

The goaltender market is full of misrepresented talent and is the perfect environment to identify market inefficiencies. Shot quality does exist in large samples. And soon, thanks to better and more camera angles, the data will be available to end the debate once and for all. With that, the inefficiencies on which most teams currently base their statistical models will disappear.

]]>
Hockey NHL sn-blog-entry
Graham Hughes/CP Tomas Plekanec, Carey Price, Tom Gilbert, Jeremy Morin, Same ugly problems continue to plague Habs Fri, 07 Nov 2014 12:22:01 EST Fri, 07 Nov 2014 16:14:19 EST Chris Boyle The Montreal Canadiens are off to a good start, but look a little deeper than their record and you’ll find some disturbing trends. Chris Boyle reports.

]]>
In an atmosphere of instant opinion, perspective often gets buried by hot takes, Twitter outrage and surface-level analysis. Nothing highlights this better than the 2014-15 Montreal Canadiens, because while the roller coaster climbs and drops, nothing has really changed outside of perception and expectations.

The Habs’ hot start coupled with a young core of stars in their prime have take expectations to an absurd level. Everywhere you turn you are being told the Habs are experiencing a historic start. Yahoo has them as 14/1 favourites to win the Cup.

This season’s team has even been compared to the 1962 Canadiens because their win totals are the same. Obviously, connecting this edition of the Habs to the dynastic teams of the past is meant simply to ratchet up the hyperbole machine. The problem is this start isn’t their best in 52 years, it isn’t even their best start in 12 months if you care to use perspective and context.

The NHL recently adopted ROW (regulation plus overtime wins) as a tiebreaker to slightly penalize teams with a lot of shootout victories. Victories that didn’t exist before 2006. When you take into account that regular season overtime didn’t exist until 1983, comparing eras is not apples to apples. The 1961-62 Montreal Canadiens started the season 7-2-1—seven regulation wins, two regulation losses and one tie after 60 minutes. The 2014-15 edition is 9-4-1, with four wins in regulation. Its lofty ranking is entirely due to two shootout wins and two overtime victories. The 1961-62 team had a plus-17 goal differential, the current squad is a minus-8.

If we are going to compare the two in any meaningful way, we need to look at regulation success, since almost 40 percent of the comparable is entirely based in regulation. I ranked every Habs team by winning percentage based on regulation success and counted everything after the third period as a tie. Goal differential worked as the tie-breaker. What I found is that the 2014-15 Montreal Canadiens’ start ranked 31st behind such notable seasons as 2007, 2009, 2011, 2013 and 2014.

habs2

(Click here for the full chart)

The reality is, they are holding script with the same quick jumps they have been receiving for the past half decade.

One of the main concerns with this team has been personnel use and tactical strategy. With the removal of possession anchor Douglas Murray, some shrewd off-season moves by GM Marc Bergevin and coach Michel Therrien’s tactical adjustments in the playoffs that saw the Canadiens crack the 50 percent Fenwick close barrier, there was optimism that the Habs were about to make the jump to legitimate Stanley Cup contender.


More NHL on Sportsnet:
Subscribe: Rogers GameCentre Live
Rogers Hometown Hockey | Broadcast Schedule
Sportsnet Fantasy Hockey Pool


Instead, the same old issues (poor possession, dump and chase, and poor zone entries on both sides of the puck, suspect defensive zone coverage, curious line deployment—ie. Dale Weise on the first line, etc.) are haunting the Habs. They’re just being temporarily obscured by come-from-behind victories and shootout dominance. It is not easy to criticize a team with a 9-4-1 record that’s on pace for a 111 points, even after going down this same road last season with the Maple Leafs, who started 10-4 last season but ultimately were buried by their poor underlying numbers.

The Canadiens goal differential is a disaster, but it is mostly due to some meltdown games versus the Flames and Lightning. One of the goals of the Shot Quality Project was to isolate team play from the highly erratic game-to-game performance registered by goaltenders and shooters. It is the same philosophy as possession metrics, but offers up some extra information based on the quality/distance of shots registered.

When viewed as a whole, it appears disastrous, but when viewed on a game-to-game basis the Canadiens aren’t that absurdly off where they should be. Through expected-goal totals, I registered a win if their expected total was over .5, a loss if under .5 and a tie if the total was within .5 of their opponent. For actual results I removed the overtime results.

habs3

The problem is that a .500 record is nowhere near a Stanley Cup contender. At issue is a brutal defence that will bury the Habs if it continues. The Canadiens have an elite goaltender in Carey Price who manages to mask some of his team’s deficiencies and although his .907 looks like a problem area, it is actually much higher than his expected total of .895 based on his 2014-15 workload.

[polldaddy poll=8432269]

Over Michel Therrien’s two-plus seasons in Montreal, the opponents expected shooting percentage is close to 9 percent, not the current 10 percent. With the presence of Price, if that rises to even slightly below average the goals against should normalize.

When actual production is compared against expected production, it becomes clear that goal scoring has been the biggest concern.

Habs_For_Vs

Over their first 14 games, the Canadiens are producing nine goals fewer than expected based on their shot production (location + movement). You can see above that Montreal’s nose dive began around the time of their western road swing where they lost a winnable game to the Oilers. With their offensive talent, they should rebound to a point where they don’t need to rely on stealing bonus points to maintain their high seed. The issue, as always, remains Therrien’s tactics, which have worked as a possession anchor his entire career.

Bergevin has divested the team of most of the negative possession players, but the Canadiens continue to surrender the blueline, dump pucks in on the attack and elevate players like Weise to the top line. (Three of their five worst games—expected wise—have come with Weise on the top line. Needless to say, it’s bad idea to add a brutal possession player—38% Fenwick Close—to the top line. Max Pacioretty and David Desharnais went from 58 and 60% Fenwick Close players to 34% players with Weise.)

As long as the Canadiens stay below the .500 possession line, they can’t really be taken seriously as a Stanley Cup contender. But as their shooting percentage moves closer to the expected result, their defensive coverage gives Price and Dustin Tokarski a better chance to succeed and their power play starts clicking, the Canadiens should continue to maintain their status as a dangerous team one tier below the true contenders.

]]>
Hockey NHL MON sn-blog-entry
Walter Iooss Jr./Getty dryden_ken_canadiens Stats say the greatest NHL goalie is… ? Fri, 24 Oct 2014 13:58:17 EDT Mon, 17 Aug 2015 08:32:29 EDT Chris Boyle With new statistics available, Chris Boyle determines who stands alone as the NHL’s all-time best goaltender.

]]>
Editor’s Note: Statistics for Henrik Lundqvist in this article do not include the 2014-15 season.

When the Shot Quality Project debuted on Sportsnet.ca during the 2013–14 NHL season, it turned heads in the hockey community with both the size of the undertaking and the fact that it undercut traditional stats.

Since then, writer Chris Boyle has spent hundreds of hours tracking thousands of shots to break new ground on how best to analyze and grade goaltenders. When he first got his hands on some of the information detailed below, he casually kicked around the idea of coming up with a formula to compare goalies across eras. We were, to say the least, intrigued.

One of the difficulties in any assessment of the greatest goaltenders ever is the lack of save-percentage data prior to the 1980s. Rankings involving goalies from previous decades have relied on team results and reputations forged by storytellers. But two books took on the task of filling the gaps by collecting old data from historical scoresheets to determine save percentages. The Hockey Compendium: NHL Facts, Stats, and Stories fills in data from 1954–55 to the final years of the Original Six era, and Goaltenders: The Expansion Years (1967–1979) takes care of the rest.

Armed with that data, I created league averages from 1954–55 onward and adjusted for individual eras. That allowed me to come up with a baseline season for a league-average goalie: 25-25-10, .898, 3.02. With that, I had a level playing field on which to measure 59 years of NHL netminders.

I weighted my research toward save percentage, but also accounted for on-ice success, longevity and innovation. Initial research had Ken Dryden so far ahead of the field that he was Gretzkyesque. I evened the statistical playing field by comparing the eight-year (the length of Dryden’s career) statistical primes of each goaltender. I also adjusted for shootout victories, which have inflated win totals over the past decade.

1. Ken Dryden

(Era adjusted, eight-year prime averages, based on a 60-game workload) The biggest knock on Ken Dryden has always been that his record was the result of playing behind a great team. One way to expose team effect is to assess all the goaltenders on a given squad. Reigning Vezina Trophy winner Tuukka Rask’s career save percentage is .003 better than that of his backups. Dryden was .019 percentage points better than his backups during his eight-year career, putting an end to any Chris Osgood–level criticism.

Dryden produced an astronomical save percentage during his career—.024 above league average. To put that into perspective, in 2013–14 numbers, that’s a .938 on average (Rask’s was .930 last season). During the 1975–76 season, Dryden registered a .927 when the league average was .890, the equivalent to .945 last season.

Couple the stats with on-ice results and it’s clear that Dryden is, without a doubt, the greatest goalie of all time.

2. Patrick Roy

Both Patrick Roy and Dominik Hasek were contenders for the No. 1 spot but statistically couldn’t match Dryden’s dominance. Instead, they battled to be the best of their generation.

They registered almost identical numbers, and while Roy played on better teams during his prime, he still produced .020 above his backups, slightly trailing Hasek’s .025. When comparing their on-ice accomplishments, Hasek outdistanced Roy in individual awards, but Roy’s playoff accomplishments cancel out Hasek’s Harts. In effect, during their primes, they were statistical equals.

But Roy gets the nod by a razor-thin margin. The defining factor was innovation. He is godfather to the current generation of goalies — The Windows 1.0 to today’s 8.1 version.

His technical innovations led to the equipment alterations and new skating techniques that have made goaltenders more dominant today than ever.

3. Dominik Hasek

The Buffalo Sabres are retiring the Dominator’s number this January, a well-deserved tribute for one of the greatest to ever play the position.

4. Jaques Plante

Plante does not match the statistical greatness of the top three, but, like Roy, he had a major impact on the game with technical and equipment innovations.

He will always be known for being the first to regularly wear a mask, but he also changed the way the game was played with his propensity for leaving the net, playing the puck and communicating with defenceman being pressured on the forecheck. Plante also registered the greatest save percentage in history with a .944 (.041 above league average) in 40 games with the Leafs as a 42-year-old in 1970–71.

5. Bernie Parent

The argument can be made that, at his best, Bernie Parent was the greatest goalie ever.

During the 1973–74 season (which didn’t have overtime), Parent had a record of 47-13-12 and a .933 save percentage against a league average of .896. With the Stanley Cup and Conn Smythe Trophy included, it’s the most impressive campaign by any goalie in 59 seasons.

He followed that up with a 44-14-10 record and a .918 save percentage (.028 above average) in 1974–75, adding another Cup and Conn Smythe for good measure.

[polldaddy poll=8393450]

6. Tony Esposito

“Tony O” dominated the early ’70s. He was the undisputed goaltending champion of 1970 and ’71 before getting overshadowed by Parent and Dryden. Esposito also helped popularize the butterfly and made equipment tweaks that advanced the position.

7. Glenn Hall

There’s a big statistical gap between Hall and the top 6, but he originated the butterfly.

Plus, his numbers must have suffered from the grind of playing every single game for seven straight seasons. That’s an absurd accomplishment.

8. Martin Brodeur

Brodeur is considered the best goalie ever by some thanks to longevity. But his statistical prime doesn’t hold up to his peers’ (Roy and Hasek) and his most impressive stats (wins, shutouts) are heavily team-influenced.

9. Ed Belfour

During their primes, Belfour and Brodeur were almost statistical equals. Belfour was better in his later seasons, but his late NHL debut, at 23, cost him a spot or two.

10. Henrik Lundqvist

This final spot was a titanic struggle. Billy Smith lost out because Chico Resch and Roland Melanson essentially matched his numbers from 1975–84, while a late start to Johnny Bower’s career hurt him. I’ve watched the overlapping primes of Lundqvist and Roberto Luongo and, though they’re nearly statistical equals, “the King” takes it.

]]>
Hockey NHL MON sn-article
Chris Young/CP Jonathan Bernier Jonathan Bernier; Toronto Maple Leafs Why Bernier is Maple Leafs’ best bet in goal Mon, 29 Sep 2014 13:20:33 EDT Tue, 30 Sep 2014 11:17:58 EDT Chris Boyle Despite what some fans in Leafs Nation might be holding on to, Chris Boyle has no doubt that Jonathan Bernier is a better goaltender than James Reimer. And he’s got a whole new way to prove it.

]]>
Analyzing goaltenders and discovering any pattern of repeatability has proven difficult for even the best of the analytics community. There have been problems with the data accumulation leading to flawed data sets, sample-size issues and struggles isolating defensive impact on a goaltenders numbers.

I ran into these sample-size issues with James Reimer when I wrote last season about Olympic goaltender contenders last December. My self appointed window focused on performance heading into the Sochi, a window of about 60 games. The time frame highlighted the best hockey of Reimer’s pro career and an assessment, about 30 percent of his career total, leading to a favourable assessment.

With a goaltender controversy always looming beneath the surface in a market like Toronto, I wanted to look into the entire careers of Jonathan Bernier and Reimer to make a better assessment. When added to traditional scouting methods, which favour Bernier’s efficiency, it became pretty clear that Bernier is the better goaltender.

As my data set has grown, I’ve been afforded better information to create expected save percentages. My expected save percentages are now tailored to individual goaltenders by assessing expected outcomes from each shooting position on the ice. That allows me to calculate what an average goaltender outcome would be based on the same workload faced by the goaltender I am studying.

Last season I assessed Bernier and Reimer based on the same average. This season I can create an average based on their actual workloads. Bernier’s expected save percentage for his career is .917 whereas Reimer’s is .920. Much higher than the league average of .913.

Reimer_Numbers-copy

I mapped Reimer against the results of the average based on my 50,000-shot sample. Yellow indicates what an average goaltender should produce in Reimer’s individual circumstances; blue represents what Reimer actually produced—below average in every major category. His performance on clean shots is a concern because they make up the largest sample a goaltender faces. This could be a function of getting into position slightly late. That can spiral because you can’t read and assess before the shot, your feet are set late, your hands trail the play and you begin to react instead of acting based on a plan and reading the ice.

Reimer’s lateral struggles don’t surprise me. I find he has issues with balance, which result in him over-committing and finding himself on all fours quite frequently. His hand positioning also doesn’t allow him to control shots with his stick in the same manner as Bernier, which is why rebounds come off his pads allowing for less control.

I then mapped Reimer versus the expected average from each location on the ice. The areas marked in green are where he performed above the average expectation, red is below average, white is the league average.

bernierreimersave

This is where we begin to see Bernier really separating himself from Reimer. Goaltenders generally are vulnerable in the home plate area, but Bernier excels in a lot of areas where Reimer struggles. Bernier is also a better skater than Reimer, which allows him to beat plays. Bernier’s reads and skating allow him to get to where he needs to be before danger arises on plays high in the zone. Add in better hands in better positions and he corrals pucks with his body or places them in less dangerous areas. On plays down low, superior balance and lateral mobility allow Bernier to also beat plays along the ice, hence his above-average marks from 10-12 feet in front of his net.

Bernier_Numbers

One of the reasons that “rebound control” became such a big talking point in Toronto is because it is one of the most obvious areas Bernier is markedly superior to Reimer. Bernier gives up fewer rebounds and is much better recovering to stop them. But that is just one of many small factors that highlight his overall superiority.

Bernier_Reimer_GBG

We can see from their rolling averages that Bernier was on a steady upward trajectory up until last season when he began to plateau. Upon his arrival in Toronto, he settled into a slightly above-average goaltender trending as high as +.04 last season. Outside of his pre-Olympic spike, which settled back into his slightly below average range in the second half of 2013-1, Reimer has been pretty steady over the past two seasons in the -.05 range.

This is where we run into controversy. Based on scouting reports and these numbers, the Leafs are actually nicely situated with an above-average goaltender and a competent backup who can step up in situations when needed to bridge injury gaps and allow the starter some rest. The problem is expectations. When your backup has established that he can start, a perfect situation can quickly destabilize into an unhealthy one.

Reimer is still young and he may need to make a run with another organization to prove where his ceiling lies, but the Leafs made the right choice last season. Bernier was a significant upgrade.

]]>
Hockey NHL TOR sn-blog-entry
FRANK GUNN/CP LINDROS01 Eric Lindros should be a Hall of Famer Tue, 24 Jun 2014 15:43:42 EDT Tue, 24 Jun 2014 19:15:58 EDT Chris Boyle He is undoubtedly one of the most unique players in hockey history—a trailblazer on and off the ice. And yet, Eric Lindros is not in the Hockey Hall of Fame. Chris Boyle puts his career in perspective.

]]>
For years I have wondered if Eric Lindros would eventually become a sympathetic character. He spent most of his career fighting for himself and his career, which created a lot of animosity at the time. He was a trailblazer at a time when his actions were frowned upon. That has probably played a part in him not being inducted into the Hockey Hall of Fame. I find Lindros fascinating because fans forget how he was perceived while still playing and view him instead through a 2014 lens. That is interesting because his off-ice actions are not viewed with the same selective amnesia.

Twenty-five years ago, players didn’t take control of their careers in the manner he did. Some point to Lindros’s decision to force a trade to from the Sault Ste. Marie Greyhounds to the Oshawa Generals as tampering by his meddling parents. As a parent of a 10-year-old, I don’t consider refusing to send my child five hours away from home meddling. I consider it good parenting. We also forget that when Lindros entered the NHL draft in 1991, the free-agency system of the time would have bound him to the Quebec Nordiques until he was 31. Today’s players have better options than Lindros had, but he used his considerable leverage to get a situation he found acceptable.

Our understanding of concussions did not exist when Lindros suffered from his initial brain injury. His first concussion came less than one season after Craig MacTavish played his final game without wearing a helmet. Lindros sustained seven concussions between 1998 and 2000. In the current climate, how would we react to Sidney Crosby being put in that situation repeatedly; having his career derailed by pressure to come back early, and then have it basically ended by a vicious head shot, one that was celebrated at the time but is an instant suspension today?

Lindros burned bridges in Philadelphia while seeking a specialist after what he felt were numerous examples of the team doctors putting the team’s needs above his own. Think of how the Penguins worked in conjunction with Crosby to diagnose his injury and contrast it to how Lindros’ experience unfolded.

The first half of Lindros’s career–when he was an utterly unique and dominant force of nature–has been overshadowed by the player he became after repeated head trauma and the infamous hit to the head by Scott Stevens during the 2000 playoffs. We can see the drop off in performance when illustrated against his adjusted points-per-game data (I adjusted Lindros performance to reflect his era).

Lindros_Concussions_470

Pre-concussions he was power forward with a mix of size and mobility that was never before seen in the NHL. His playing style may have been unsustainable and Lindros struggled to maintain his health before the concussions, but when healthy, he was a punishing hitter and fighter who could score at a Crosby pace. He made Team Canada at 18 years old in 1991 before playing a game in the NHL and did not look out of place. He was the NHL MVP in 1995 and was named the captain of Team Canada in 1998 by the same guy who stripped him of his captaincy in Philadelphia two years later.

That he has not made the Hall of Fame at this point is absurd. Longevity is a concern, but it wasn’t a concern for Mario Lemieux, Mike Bossy, Cam Neely, Pat Lafontaine, Pavel Bure or the recently enshrined Peter Forsberg.

I understand that offensive point totals aren’t the definition of complete hockey players. But seeing as it is essentially the standard for awarding the Hart Trophy, it is a legitimate standard to argue the merits of Lindros’s career. I adjusted the numbers to reflect eras—comparing Bernie Federko’s 100-point seasons to Forsberg, who played in the lowest-scoring era in NHL history, is not a proper reflection of his performace against his peers.

Lindros_HHOF_Career_470

Lindros is on par with Mike Bossy and Pavel Bure, and is miles ahead of his closest peer in the power-forward category, Cam Neely. Now some of these numbers are influenced by career longevity, but I didn’t want to penalize those players in the comparable so I removed post-prime seasons and charted players from the age of 18 to 30.

Lindros_HHOF_Prime_470

When comparing Lindros to Hall of Famers in their primes, he is up there with players like Brett Hull, Mike Bossy, Guy Lafleur and Marcel Dionne. Ahead of guys like Joe Sakic, Steve Yzerman and the recently inducted Mike Modano. These guys are no doubters, but Lindros has still not been deemed worthy. To illustrate how he would stack up against the current NHL elite, I added in the today’s generation of superstars for context.

Lindros_HHOF_Current_470

Lindros is not far from Alex Ovechkin and Evgeni Malkin, and that includes his post-concussion results. Before suffering his first concussion, Lindros was up in the rarified air currently occupied by Sidney Crosby, the guy who is no longer the best player in the NHL because he was eliminated in the second round of the playoffs.

This isn’t even a debate. The Hockey Hall of Fame is so easy to get into there are actually real discussions about Chris Osgood’s legitimate claim because he has 400 wins to go with a league-average save percentage. Injuries aren’t a concern. If they were, Cam Neely wouldn’t be fist pumping in the press box wearing a Hall of Fame ring. It isn’t about being a winner because Marcel Dionne, Mike Gartner, Peter Stastny and Dale Hawerchuk are all enshrined even though they never sniffed the Stanley Cup final.

Eric Lindros is a Hall of Famer. Anybody who watched him play during his prime knows this. What we are seeing is Lindros paying for his decisions to stand up to the establishment. He didn’t play ball when he pushed his rights as a player to control where he played, nor when he challenged management for putting the team above his personal health. For these reasons, his punishment is to sit and watch, and understand his place.

]]>
NHL PHI sn-blog-entry
(Mark J. Terrill/AP) lundqvist_kopitar Stats say Kopitar deserves Conn Smythe Fri, 13 Jun 2014 13:00:23 EDT Fri, 13 Jun 2014 14:43:05 EDT Chris Boyle The race for the Conn Smythe trophy twists and turns in the Spring’s sports pages. But a few key stats can help show who will claim it in the end.

]]>
Every season there seems to be great debate as to who will win the Conn Smythe trophy, with power rankings that begin after just a couple of games. That early, it’s surely premature, but if we didn’t do early ranking we would never hear the words “San Jose Sharks” and “Conn Smythe” in the same sentence.

Players jump in and out of contention and everybody has an opinion up until the Stanley Cup is finally presented. Media and fans argue, but it is generally fairly predictable who will actually be named the playoff MVP, based on historical precedent. Nine out of ten times the winner comes from the Stanley Cup–winning team. So when a team like the Kings jumps ahead 3-0 we start to peruse their roster for candidates.

Generally the search begins with the goaltender, because they have won the award 33 percent of the time. The Kings have a Conn Smythe–winning goaltender on their roster, but he has been mediocre to bad all playoffs.This is where the debate really begins this season, because over the last 20 seasons, a post-season like Quick’s registers so poorly that he is not even a realistic candidate.

Conn_Smythe_Goaltenders

Outside of Brodeur being snubbed for Claude Lemieux in 1995 (+.034), if a goaltender registers an elite save percentage during a Cup winning season, they generally walk away with the Conn Smythe. Brodeur has been snubbed twice, the other being with probable cause (Giguere +.026 in 2003), while Ed Belfour was snubbed for Nieuwendyk in 1999, a post-season where a legit argument could have been made for Hasek. If the Rangers pull of their miracle comeback Lundqvist is a lock to win the trophy.

Chris Osgood lost out in 2008 during a strong statistical run, but he was defeated by the second-most obvious Conn Smythe selection: the forward who leads the league in scoring for the Cup winning team. In that case it was Henrik Zetterberg put together a strong run for the Wings. Scorers who put up elite point per game stats are rarely overlooked unless they come head to head with the elite goaltender performance (Ward>Staal in 2006) or, of course, are outscored by a teammate (Malkin>Crosby in 2009).

Conn_Smythe_Forwards

What this means is that Anze Kopitar has jumped into contention for the Conn Smythe, not because he is having a history-making playoff run, but because he matches the historical criteria the best at this point. Justin Williams is a dark horse because of the media getting behind the Mr. Game Seven moniker, even though wingers rarely win the award. But hey, as we saw with Claude Lemieux in 1995, Scott Stevens in 2000 and Mike Vernon in 1997 reputations can occasionally take a life of their own during weak years and could result in an award on Williams mantle.

Similar to the Hart trophy, media and fans don’t really know how to assess defenders properly because they rarely dominate statistically. This results in defensemen rarely being awarded the trophy, even though an argument can be made that they are the most important players on the ice. Doughty has recently been the anointed the best player in the league recently by some, as the media has developed a schizophrenic nature of changing this individual every week.

Conn_Smythe_D

Lead the league in playoff scoring as a defenseman like Brian Leetch and you are a lock for the Smythe, but generally these players vault to the forefront only because a lack of strong options at other positions. Lidstrom won his award during a season where Hasek was league average and Yzerman barely registered a point per game. Same with Niedermayer in 2007 and Stevens in 2000. This opens the door for Doughty’s candidacy. His performance is not far off Lidstrom’s in 2002 and with no obvious winner standing out, he could be a dark horse in 2014.

The 2014 playoffs is also open for the once-in-a-decade losing team player trophy steal. If the Rangers can extend the series to 7 games, Lundqvist might be able to sneak away with the hardware in a field where the lack of leader could easily split the Los Angeles Kings vote. Of the five times the award was given to the losing team, four of those times it has been awarded to a goaltender.

Mr. Game Seven could very easily ride his “clutch” reputation to the Smythe award, but the most deserving King is Anze Kopitar. Not only does he meet the criteria of leading his team in scoring, but he has also managed to dominate traditional and #fancystat categories while facing elite competition.

Kopitar

 

Kopitar has outperformed his expected output while driving the Kings’ play. His possession numbers are strong considering the match ups he’s face during this Kings playoff run. Kopitar struggled early at 5 on 5 as the Sharks matched him up against Patrick Marleau, Logan Couture and Matt Nieto, but as the series wore on the Slovenian began to take over. When the Kings knocked the Sharks out, his assignment became Ryan Getzlaf, Corey Perry and Devante Smith-Pelly. His task against the Hawks? Jonathan Toews, Marian Hossa and Bryan Bickell. In the Finals it is more of the same.

Kopitar isn’t just negating the top of the other teams’ lineups, he is dominating those match ups. He has always been underrated playing on the West Coast, but I do find it interesting that, after defeating Jonathan Toews—the man that inherited Sidney Crosby’s best player mantle in the Conference Finals—the imaginary belt went right past Kopitar to Drew Doughty. It is possible  Doughty steals the Smythe, I can live with that, but Kopitar should win it because it is time he is recognized as a superstar.

]]>
Hockey NHL LA SJ sn-blog-entry
Mark J. Terrill/AP lundqvist_henrik Kings goalie Quick too mistake prone Thu, 05 Jun 2014 08:49:15 EDT Thu, 05 Jun 2014 16:04:00 EDT Chris Boyle In any given game — like Wednesday’s Stanley Cup Final opener — Jonathan Quick’s gambling style can win. But in the long haul, Henrik Lundqvist is the more reliable goaltender.

]]>
The 2014 Stanley Cup Final has provided us with a stark contrast in goaltending styles. The patience of Henrik Lundqvist vs the impetuous Jonathan Quick. A battle of depths as well as the philosophy of getting to the play ahead of the puck or just getting to the play to stop the puck. Outside of their athleticism, they represent two totally different philosophies of how to play the position.

History has proven that Henrik Lundqvist’s intelligent measured approach is more repeatable than the freelance approach of Jonathan Quick.

When you analyze Lundqvist’s play, you can see his hockey IQ in progress. The philosophy behind his inside out technique is based on simple math. The tighter you are to your goal line, the shorter your routes to the puck. This is extremely helpful on backside and rebound recoveries because the distance he has to travel is minimized because of his starting point. It also allows him to get into position faster and identify potential dangers earlier in the process. He isn’t tied to his goal line as he adjusts to situations when needed with added depth when he identifies the lack of lateral danger.

When we isolate both goaltenders most common trait, athleticism, we see it applied in entirely different ways. Lundqvist uses his athleticism late in the play, it allows him to play deep and maintain his feet and read and react. His style of play has proven difficult for goaltenders to replicate because of his ability to make reactionary saves while standing on his goal line.

Jonathan Quick’s athleticism is on display every moment he is on the ice and becomes readily apparent within the first 5 minutes of action. He is extremely aggressive and is comfortable in playing at extremely aggressive depths because he is confident his athleticism can bridge the longer recovery routes he always places himself in . This confidence leads to a gambling style in which he drops early on plays and explosively recovers. He plays extremely low and his flexibility allows him to seal the ice after initial recovery pushes propel him towards the puck. Elevating the puck is key to beating him, but he attempts to dominate the angle and block elevation with frontward leans.

When I watch Quick play, I feel cheated. Although he offers a fan friendly style of play, I feel if he took a more analytical approach to the position and applied his athleticism only when needed, his results would be more repeatable and he would be among the games elite. Through six seasons he hasn’t proven his current style is repeatable spending half his career at average or below.

Quick_SV-4701

Goaltenders in general tend to ebb and flow because they can’t control their environment, but Quick is extremely reliant on coverage to his blindside and second chance opportunities because of his outside/in approach to the position. It showed during the first round series with the Sharks when the Kings failed to cover for him when wandering from the blue paint.

Lundqvist plays a style that is not as system reliant. Sure he is still vulnerable to the same type of dangers as Quick, but he mitigates the damage through his inside/out style. If a team is giving up too many backdoor plays, he becomes less exposed because of the decreased distance he needs to travel. Not once during Lundqvist’s regular season career has he finished below the league average in save percentage.

Lundqvist_Quick_rev

Quick’s wild inconsistency has been front and center during the 2014 playoffs. I tracked Lundqvist and Quick in the same manner as Rick Nash last week and created an expected goal total based on their performance on the way to the 2014 Stanley Cup Finals.

Lundqvist_Quick_rev

We can see by his expected save percentage that Quick was exposed to a slightly tougher workload than Lundqvist overall (.914 to .918), but his inconsistent nature almost cost the Kings in the first round as they failed to provide the adequate support for Quick’s style. Contrast him to Lundqvist who has maintained a consistent level above expected results for the majority of the playoffs. It is easy to see how it is rational to believe that a more structure based goaltender could benefit a strong team like the Kings. If we look at the results of Quick’s less aggressive teammates during Sutter’s tenure we get some head scratching numbers from unexpected sources. Ben Scrivens (+.017), Jonathan Bernier (+.010) and Martin Jones (+.020) all benefitted from less is more over the last two seasons.

Quick always remains a wildcard for me because of his style, but over small bursts he can provide elite goaltending to rival Lundqvist. Is Lundqvist the better goaltender? Without a doubt. His data backs it up, as does his approach to the game. He is the biggest reason the Rangers are playing for the Stanley Cup. The problem is these playoffs have shown that the Rangers need an elite level Lundqvist to win playoff games, the Kings have been doing it against the elite of the Western Conference with backup level support. Lundqvist is capable of stealing the Final, but the Kings might not require elite level goaltending to win again.

]]>
Hockey NHL sn-blog-entry
Nick Wass/AP NASH01 Rick Nash producing offence, if not goals Thu, 29 May 2014 13:28:40 EDT Thu, 29 May 2014 14:38:15 EDT Chris Boyle The New York Rangers’ highest-paid player has been criticized for a lack of production. But look deeper and you’ll see he’s creating chances, but not getting bounces.

]]>
It’s hard to avoid criticism during the playoffs when you aren’t scoring and you make millions per season. Fans and media are reactionary and when guys like Sidney Crosby and Rick Nash aren’t hitting the score sheet, they are immediately identified and buried. We love to measure character level and label winners and losers based on tiny pieces of information.

Jonathan Toews entered the 2014 playoffs with six goals in his previous 36 post-season games. Stories of his frustration were rampant when the Hawks almost lost to the Red Wings in 2013, but with eight goals in these playoffs, the previous three seasons of frustration disappeared. Toews has re-emerged as Captain Clutch, the true greatest player in the NHL.

The problem is slumps are generally sample-size based and when we cut them into smaller segments because of playoff years, the natural inconsistency leads to wild performance fluctuations. It is why the analytics community relies on shot metrics, because they offer a better look at a player and remove a lot of the noise.

If we look at Toews’s playoff production early in his career (2008-09 and ’09-10) and extrapolate and 82-game campaign, he is a 30-goal Conn Smythe winner. From ’10-11 through ’12-13, though, he dips to a disappointing 14-goal player. This year he’s up to 47 goals because he is shooting 26 percent. And if we look at his entire 92-game playoff sample, he is the same 25-30 goal guy he’s been over close to 500 regular-season games.

Unfortunately for Rick Nash, with an $8-million salary comes the demand of goals. It is of no interest to fans that Nash has been a fairly dominant possession player and his team is winning the battle while he is on the ice. Or that he has accomplished this while managing tough assignments during the 2014 playoffs. He isn’t scoring and this is a problem, a problem for which we are dying to define him until the narrative changes with a hot streak.

During my research, I identified that, regardless of location, if we added in slightly more context the numbers could shift. Shots preceded by passes, rebounds or pucks that are tipped shift the odds to greater success. We tend to look at the game in simple concepts. When players disappoint us by missing scoring chances too often, we often vilify them for their failure. The reality is, even on the greatest scoring chances the odds of success rarely move above 50 percent.

Nash_Tip

The shot quality project isolates areas on the ice that have higher scoring probabilities, but even the one pictured above where Nash tips a puck from the lip of the crease in front of Marc-Andre Fleury only results in a goal four of 10 times. During a playoff series, you could conceivably tip six pucks and come up empty, even though statistically you should have produced almost 2.5 goals. This is the danger of judging a small sample.

One of the things I’ve been working towards is a goal-expectancy chart. It’s still in its infant stages because I have just over 40,000 shots, but it’s accuracy will improve with a larger sample size. It’s large enough that I can get a good gauge of shot quality. So with the noise surrounding Nash, I decided to put his production to the goal-expectancy test.

I reviewed every shot Nash took during the 2013-14 season (regular season and playoffs up to Game 3 of the Canadiens series) and calculated his expected goal total. Nash has been accused of weak perimeter play, and of the 320 shots I looked at, 75 registered at 1 percent or less of scoring. These included dump-ins from the neutral zone, clean shots from just inside the blueline and shots from along the boards from an impossible angle. These type of shots create noise in our general shooting-percentage numbers and lead to explanations like luck—a divisive term with fans because it tends to take credit away from the player. “Luck” is used because it simplifies the concept instead of other terms that fans tend to reject, like probability and regression.

The more shots you produce, the more fortune you will find and if Nash continues to fire pucks from “good ice” pucks will find ways into the net. The more you are in that position, the better the odds that players will make mistakes or bounces will find their way.

I tracked his results game-by-game and charted his actual production (goals) versus his expected goal total.

Goals_Expected

There is a pretty straight progression line (in blue), a stark contrast to the streaky pattern most goal-scorers produce. Pretty much game-in and game-out, Nash creates scoring chances and drives play. That is reflected in the consistency of his expected goal total—he just isn’t consistently rewarded with goals. Marian Hossa had a brilliant response earlier in the playoffs when he condescendingly said that the difference in the perception surrounding him was the result of him scoring a goal in that night’s game. His process hasn’t changed, just the way we absorbed the result.

Nash more or less maintained a steady goal-scoring pattern until early January when his production took off with 11 goals in 15 games (in red). He maintained a consistent record of producing scoring chances, but the expected numbers trailed his actual production until he flat lined in the playoffs and his numbers dipped below expectation. He has recently begun to produce again with three goals in the first five games against Montreal.

This measured approach isn’t as fun as defining a player’s character or speculating about injuries—especially if you enjoy the teeter totter approach of that game. Nash will never escape criticism during these droughts because of the expectations that come with his pay cheque, that is the reality of the situation. But he is a 35-goal guy who played at a 30-goal pace during the season and had a regression kick in at the wrong time of year—the time of year when legacies can be defined. Fortunately for Nash, our memories are short and two or three timely goals over the next couple of weeks will give us all collective amnesia as we scramble to write a new reputation for him.

]]>
Hockey NHL NYR sn-blog-entry
Ringo H.W. Chiu/AP Dustin Tokarski, Brian Gionta Tokarski the smart move for Therrien Thu, 22 May 2014 16:32:47 EDT Thu, 22 May 2014 16:53:42 EDT Chris Boyle Going with an untested rookie in goal during the playoffs doesn’t always work out, but if your only other option is Peter Budaj, you really can’t lose.

]]>
One of the dominant story lines entering the Eastern Conference final was the goaltending matchup. Although the Bruins series also featured elite goaltending, the shot quality project had questions about Rask’s greatness, questions no one has through close to 4,000 shots with Henrik Lundqvist.

Although the Western Conference final is considered by many to be the true Stanley Cup final, any team backed by a Lundqvist or a Price would have a puncher’s chance. Of all the goaltenders I have reviewed, they top the list.

Stylistically, Lundqvist takes a deeper approach to the position, something that I delved into in February. That allows for shorter routes to pucks and easier recoveries on backdoor attempts. We can see this in the data sets as Lundqvist maintains a significantly higher save percentage on lateral feeds. Price’s more conservative depth makes him more effective on clean looks and his outstanding footwork along the ice made him more effective on rebounds. Essentially they are equals against a league-average workload; Lundqvist edged Price out by .922 to .921.

Price_Lundqvist

It was an amazing matchup, almost like an arms race. The Canadiens were able to match the Rangers in goal, so the outcome would be slanted by the skaters. Then Chris Kreider recklessly (not intentionally) put the Canadiens’ Cup hopes on life support.

With Price gone and no real replacement available, the Rangers went from a slight to overwhelming favourite. Although the Habs fan base will complain about bad luck, it is generally done with a short memory. It wasn’t even five years ago when the Canadiens PDOed their way to a conference final on the back of the unheralded Jaroslav Halak. I personally look at it is the interest charge on the 10 overtime wins during the 1993 playoffs.

Although Kelly Hrudey offered criticism about starting Tokarski and others in the media second guessed it after it failed, the decision was rational and understandable. Michel Therrien has twice gone to Peter Budaj in the past two playoffs following Price injuries and he delivered 29 saves on 39 shots. A small sample, but when measured against the large sample that represents his career, Budaj fails to be even mediocre, delivering one season out of nine above league average.

Even bad goaltenders can get hot, but historically ones as poor as Budaj do not get hot enough to deliver four playoff wins. Knowing how dire the situation was I decided to search history for comparables to see if Therrien had made a drastic mistake. He didn’t.

Budaj

I used saves above average (GSAA) to decipher success rates. That is, “the goals this goalie prevented given his save percentage and shots faced vs. the league average save percentage on the same number of shots. Min. 4 shots faced per team game needed to qualify.”

I found 14 goaltenders matching this criteria (including Budaj himself in 2013). Only four managed to win more than four games and as a whole they were .500 with a league-average save percentage. Hayward made the list because of a disastrous 1986 season in Winnipeg. He would go on to register a plus-36 over the next three seasons. The miracle long shot goaltending of Alain Chevrier during the 1989 playoffs represented the ceiling for Budaj-level performance with nine of the 14 failing to match league-average performance in save percentage or more than two wins.

So Therrien was tasked with either tucking his tail between his legs and starting the guy who failed miserably in 2013 or playing on the Canadiens’ history of magical miracle rookie goaltending and offering the hope of the unknown, Dustin Tokarski.

Tokarski

The unknown, the raw rookie-backup has made plenty of appearances since 1984 (this list doesn’t include starters, guys like Patrick Roy and Price) and has registered more success than the worn, mediocre backup.

Cam Ward replaced Martin Gerber and won a Conn Smythe Trophy. Mike Vernon took the Flames to the Cup final following two seasons of Rejean Lemelin as a Vezina finalist. Braden Holtby managed to win seven games just two seasons ago in Washington. The Canadiens have used this trick many times, so it shouldn’t be a surprise that this list contains Steve Penney and his miracle 1984 run to the Conference final as well as Jose Theodore in 1997 and Halak, who came up short against the Flyers in 2010. The untested rookie backups went 70-58 and were plus-.09 greater than league average.

Franchises traffic in hope and although not many people are going to review the past 30 years of playoffs, the Canadiens’ history is littered with rolling the dice on rookie goaltenders in the playoffs. The fans are fully aware of Ken Dryden, Penney and Roy. Toss in the seamless job of mentioning that Tokarski is a “winner guy” (Telus Cup, Memorial Cup, world juniors and Calder Cup, something Hockey Night in Canada repeated 30 times) and glossing over the fact he’s an undersized mid-tier prospect who through 6,500 professional shots has registered essentially league average results (.911 to .912 save percentage), and hope is temporarily restored. Therrien played his hand brilliantly and even if Montreal is alive only for another four or five days, it was more than what existed after the news of Price’s injury hit Twitter.

There is historical precedence here, but when such stories are written, they generally don’t include the league’s best goaltender standing in the other crease. “The King” is a nickname Lundqvist has earned, the data shows that he is no paper lion. If the Canadiens can replicate their game-two performance and outplay and outpossess a Rangers team that dominates them in that category, Tokarski won’t even have to match Lundqvist. Then maybe the city of Montreal will need that riot-police budget in a couple of weeks after all.

]]>
Hockey NHL sn-blog-entry