Price / Halak - The Stats - November 2009

What a difference a month makes.

Thirty days ago the Canadiens were in the middle of a goaltending controversy. Jaroslav Halak's agent was NITWITTERING Carey Price's won/loss record, the fanbase had placed the bus in reverse and were attempting to leave a fresh set of treadmarks on Price's back and the Canadiens were plodding through their schedule at .500.

Outside of the Canadiens still flirting with the .500 mark, the attitude regarding the goaltending situation has changed dramatically. Since this article is a retrospective on November, it is heavily weighted towards Price. Halak's one start didn't offer enough of a sample to legitimately track his monthly progress. What began as an even split in October, turned into a clear separation as Price started 11 of 12 games in November, almost rendering the title of this article, and its point, moot for at least the time being. A solid performance by Halak in Ottawa suggest that the busy month of December may tell a different story.

Surprised by the turn this has taken, maybe you shouldn't be. This has been typical of the story that Price has written over the first two and a half seasons of his career. Somehow, instead of the fanbase observing the roller coaster ride from afar, recognizing that as the roller coaster descends, a peak exists for it to climb, they jump on the roller coaster eyes closed screaming for dear life on its mad descents only to extol the virtues of the ride when they disembark.

Should Price struggle in February again, it is likely that the fresh imagery that exists in everybody's minds right now will once again dissipate in a mad panic

Panic in Montreal is generally followed by the rewriting OR ignoring of history. Even with Price's strong November this controversy is not over, it will continue to bubble just below the surface. Should Price struggle in January, the detractors who are silent right now will re-emerge with the same arguments, their memories will once again be wiped clean. While the starting goaltender in Montreal is one of the most heavily scrutinized positions in sports, the easiest may be the backup goaltender in Montreal. The backup's flaws are always concealed under the their baseball caps, their past transgressions completely ignored. This will never change.

At Eyes On The Prize, we continue try to look past the surface level blame and analyze what is really going on. This month we will compare Price's performance in November to October and also look back and try to clarify the factors that contribute to a team based stat like a win / loss record and how it relates to Price and Halak.

One of the main arguments against Price in October was his inability to win games. With Halak sporting a 5-2 record and Price at 2-5, the most simplistic argument was "Price doesn't make the big saves when he needs to" and "Halak wins games". That argument ignored plenty of factors, most importantly that wins are a team based stat.

While perusing the internet I came across an interesting perspective at Brodeur is a Fraud. While trying to debunk the myth of Brodeur's career, he came up with an interesting statistic that he refers to as Win Threshold. It essentially projects out what a goaltender would have had to do to earn a point (.500) by including the teams performance in front of him. In a 3-1 victory, the assumption would be that to earn a point the goaltender would have to allow no more than 3 goals. Inversely, in a 3-1 loss the goaltender would need to have allowed no more than 1 goal.

We can calculate what I'll call the "win threshold" for the goalies on each team by taking (shots against - goals for) / shots against. This gives us the save percentage that would result in the team ending up with an equal number of goals for and goals against over the course of the season. If the goalie's save percentage is above that number, the team is likely to win more than the lose, while anything below the threshold means that the team should end up sub -.500 (or sub -.550 in the shootout era).

This simple statistic beautifully illustrates the career of Chris Osgood AND his playoff run that allowed the media to begin the "Chris Osgood is a Hall of Famer" nonsense.

In 2008-09, Detroit's win threshold was .873, which was the lowest in the league. The Islanders' win threshold was .928, which was not only the highest mark in the league but also the highest of any team since the lockout......Naturally, comparing win totals on goalies playing on the Islanders to goalies playing on the Red Wings is completely senseless.

Using this as a starting point, I compared how this statistic affected goalies on the same team. Was the Canadiens offense responsible for Halak's inflated win total, or was it his superior goaltending? What would Price have had to do to compensate for the woeful offense in front of him? Was it that Halak was receiving an easier ride?

(Let me preface this by stating that this is judging wins and wins alone. If a goaltender was fantastic and had a .960 SV%, but only needed to produce a .895 SV%, it is not devaluing his performance, only stating that his margin for error was very high. I also understand that early saves can keep a team in a period. Teams play to the score and loosen up when they are up big and that some goals are scored late in a game and alter the margin for the threshold. It is obvious that providing replacement value will never replicate the actual performance that history has provided. This theory is being used only to illustrate that wins are a team statistic and that they are heavily influenced by not only what you save, but what offense your team provides).

To put this in perspective, the league average in save percentage is .910.

In order for Price to gain the majority of his wins, he has had to offer above average goaltending.
His margin for error was slim and when he failed to balance on the high wire, he was open to criticism.
A .915 save percentage is currently good to place 20th in the league.

Halak on the other hand was allowed to navigate with a greater margin of error. He only needed to offer a save percentage of .865 to register his 5 victories. Vesa Toskala's save percentage on the season is .863. So any goaltender slightly more competent than Toskala MAY have registered a similar number of victories as Halak.

Both goaltenders would have had to provide Hall of Fame type performances to avoid the majority of their losses.

How would those statistics translate if you offered up average replacement value? If Jimmy Howard and his .910 SV% replaced Price and Halak, the Canadiens would be in deep trouble. The Canadiens would likely be in the hunt for Taylor Hall, not talking about the importance of their December schedule.

Looking at Price's losses below, it is unlikely that an average goaltender makes any impact on his losses, hell, it's unlikely that Roberto Luongo would have made much of a difference. 

Price's threshold in his 11 losses is a SV% of 0.944 and a GAA of 1.69. Numbers that would place him above Ryan Miller as the number one goaltender in the league. Six of those starts would have required a .940+ SV% just to procure a shootout. If the team continues to struggle offensively to this capacity, then it is going to require Jose Theodore 2002 MVP type performances to drag them into the playoffs. With Price and Halak entering their RFA season, THAT is a scary and expensive thought.

You would think with Halak's lack of activity in November, that his stat line would remain unchanged, but with only an 8 game sample, one strong start against the Flames resulted in an 11 point improvement on his save percentage and lopping 1/5th of a goal off his goals against average. In October, Price statistically lagged behind in all the major categories, but with a strong month has lessened the paper gap between himself and Halak. I made the argument in October that Price had a higher ceiling, but was prone to wild swings of inconsistency and the 2009 season has proven that out so far.

Through two months, what has been proven, is that the Canadiens goaltending is in good hands. With the transition of turning over half the team, and the numerous injuries to key personnel, the fact the team came through November at .500 is miraculous. At this moment, Halak's consistency offers the perfect compliment to Price, but due to the idiotic actions of his agent (Halak, Halak it a lot?), it is clear that Halak is eyeing a starting assignment, and will not likely find that in Montreal. You may be reading "Price / Sanford: the numbers" by March December. (note: as I am writing this, Sportsnet has reported that Halak has been placed on the trading block).

The fact that I can even mention a 2002 Theodore-like season is based on the strong November showing by Price. Playing behind an injury ravaged defense that produced roughly the same amount of shots as the previous month, from roughly the same areas, Price's numbers improved across the board. Price also managed to avoid a major meltdown like the Vancouver game, as well as producing a couple of 40+ shot gems against the Bruins and Predators.

With Halak pushing Price for the starting job in October, by all accounts Price responded to the challenge with hard work, which resulted in plenty of positive signs over the last month. With hard work Price's confidence level soared and his return to a hybrid style has allowed his athleticism to re-surface, because of this, Price managed to produce some miraculous saves during November. His patience has also been on display, particularly in shootouts and breakaways where once again he is using his huge frame to his advantage and refusing to commit before the shooter. With his gap control improving, he has been extremely hard to beat in 1 on 1 situations. This is a huge improvement on the tentative approach he used in late 2008-09. Add in his aggressive puck handling and his tenacity in battling to track the puck and the hope in Montreal is that Price has begun turning the corner.

Price's save percentage's also improved in almost every category in November, with his biggest improvement coming in shorthanded situations where his save percentage improved by over .200. His woefully low special teams percentage in October masked the fact that he was producing an elite level even strength percentage. With an improvement to mediocre powerplay standards, Price's statistics all experienced a significant jump across the board.

The question remains, what came first, the chicken or the egg? Did the Canadiens penalty kill improve in November because of Price, or did Price's numbers improve because of the penalty kill? Halak did not suffer the same special teams meltdown. Could that have had anything to do with the power of the opposition, or his own play?

With so many new faces and the loss of Markov, it is no surprise that the PK got off to a slow start. With the return of O'Byrne and Gill and the emergence of White and Pyatt, the October implosion by Price should be avoidable moving forward.

Interesting to note if Price had managed to produce an .878 powerplay SV% in October, his overall SV% would have improved from .889 to .920.

Price and Halak continue to get bombarded by shots and the Canadiens have once again regressed into becoming over reliant on goaltending to win hockey games. With the early season injury to the Canadiens best player (Markov in 2010, Koivu in 2002) and the continual 30+ shot barrages, this season is becoming eerily reminiscent of the 2002 season when Theodore needed to provide the difference every night.

With Price's numbers improving in November, the assumption would be that the Canadiens settled down and began to limit shots and scoring opportunities. The Canadiens actually surrendered an extra two shot attempts per game. Price still is averaging over a shot attempt against per minute on the ice, a trend that will either lead to a statistical regression or an MVP award. 

In 8 of his 11 starts in November, Price faced 30+ shots, including a team record tying 53 saves against the Predators and a 42 save gem that was seconds short of his 1st shutout of the season. Only once in his 11 games did he have less than 50 shots directed at him. Price continued to excel on shots outside of 20+ feet as he only allowed 7 of 257 shots to beat him. His SV% below 20' feet drastically improved and is pushing towards the league average. 

Price offered up plenty of spectacular saves in November, but also had some concentration lapses that resulted in some questionable goals, something that had been absent from his game in October (only 5 of 25 came in as questionable, with zero registering at fault). Although his statistics drastically improved over the month of November, with the Canadiens lack of scoring ability (25th in the NHL) and his unbelievably tough win threshold (see above), his margin for error was slim. This can be witnessed by his 0-4 record in the 4 games I credited a bad goal. In two of the games (Detroit and Atlanta) the margin of victory was one goal, the other two (Tampa Bay and Pittsburgh) were decided by two goals. As he matures, these goals should occur less frequently allowing him to become even more of a game changing player. He isn't yet ready to carry a team, but it seems the Canadiens have given him little to no choice in the matter presently.

Looking at the zones where Price has given up goals, his glove hand continues to be much maligned even though he has only been beaten high glove side 11 times through November (11/53). The reason for the questions lies in the amount of bad goals high glove side.  Out of the 12 goals categorized as suspect or bad, 6 have been high glove side, making it much easier to understand AND defend the criticism.

Looking at his November chart also gives the impression that he is very vulnerable low stick side, but looking at the video below shows that a lot of those goals came as the result of cross ice passes in which Price was not able to square up and were the result of him in transition. A small concern for me through two months has been his vulnerability through the 5-hole, an area he has not been able to cover with his stick either in transition or static and squared up.

Goals deemed questionable or bad are marked in bold.

26. Rich Peverley feeds it in front and it deflects off Bryan Little by a moving Price beating him 5-hole. 10 ft.
27. Pass from behind the net to a wide open Peverley in the slot, snaps a shot by Price low-stick side. 14 ft
28. Point shot by Mark Popovic that knuckles by Carey Price beating him high glove side. 59 ft.
29. 5 on 4 PP. Cross crease feed to a wide-open Pavel Kubina who directs it pass Price into an empty net. 5 ft.
30. Kane cross crease pass to a wide open Colby Armstrong who beats Price low stick side. 8 ft.
31. 6 on 5. Point shot. Price makes the initial stop, Patrice Bergeron rebound beats Price low-stick side. 5 ft.
32. 5 on 4 PP. Scramble. Price makes 2 saves but cannot stop Ryan Malone as he beats him five-hole. 11 ft.
33. Price, big rebound and Alex Tanguay alone, chips it over Price beating him high glove side. 14 ft.
34. 2 on 1. Scramble. Wright beats Gorges to the puck and slides it into an empty net. 6 ft.
35. Price highsticked in the throat by Mara falls down leaving an open net. Robert Lang puts it top shelf. 15 ft.
36. Point shot wide deflects into an open net off Bissonnette's skate beating Price low glove side. 12 ft.
37. Price makes initial save, Steve Sullivan all alone finishes off his rebound beating Price 5-hole. 7 ft
38. 5 on 3 PP. Point shot stopped, Sullivan at the side of the net beats Price low stick side. 14 ft.
39. Odd man rush, cross ice feed to the slot. Brandon Sutter chips it over Price, high glove side. 14 ft.
40. 3 one 2, cross ice feed one timed by Jussi Jokinen beats Price low glove side. 32 ft.                                
41. Eric Fehr snaps a shot from the face-off dot beating Price high-glove side. 25 ft.                                
42. Morrison takes a pass in front and moves opposite to Price's momentum beating him low glove side. 6 ft.
43. 5 on 4 PP. Brad Stuart point shot handcuffs Price and trickles in. 55 ft.                                
44. 5 on 3 PP. Brian Rafalski point pass tipped home by Pavel Datsyuk at the side of the net. 12 ft.                         45. Nash behind the net finds Vermette who one-times it past Price beating him high blocker side. 11 ft.
46. 5 on 4 PP. Rick Nash centers it, Stralman cheating off the point beats Price low blocker side. 24 ft.
47. 5 on 4 PP. Antoine Vermette's shot stopped, rebound back to him, beats Price low blocker side. 20 ft.   
48. Sidney Crosby snap shot from the top of the faceoff circle beating Price high glove side. 38 ft.                        
49. Bill Guerin from the corner. Price makes initial save, doesn't hold the post as he beats him 5-hole. 8 ft.
50. Crosby with a beautiful pass to a streaking Sergei Gonchar who beats Price 5-hole all alone. 8 ft.                
51. Point shot deflected in front by Brooks Laich beats Price low blocker side. 18 ft.                                
52. Alex Ovechkin one timer from the top of the faceoff circle beats Price low blocker side. 38 ft.                
53. 6 on 4 PP. Price stops a one-timer by Ovechkin, Fehr slides home the rebound beating Price 5-hole. 10 ft.

The November verdict: You have to be encouraged by the performance of Price and Halak through November and the improved penalty kill is a positive sign and a factor in Price's improved numbers. The volume of shots and scoring opportunities remain a major concern.  In October I preached patience because of the integration of 10-12 players and the introduction of a new system. With injuries creating a revolving door of players it is likely that this team has still not properly gelled. The Canadiens have fallen into a pattern of struggling to produce offense and relying on Price and Halak to bail them out, and this is a dangerous game to play with developing goaltenders. They have been able to provide some fantastic results through November, but if this continues they will struggle to make the playoffs. To provide room for improvement moving forward, the goaltenders get a B+ for November.

To view the October Price / Halak analysis, click here.