|
How bad is Craig Anderson? How…bad?
Are you kidding me?
Most likely, you haven’t read a negative word about Craig Anderson all season long––until now––as the former Florida backup boasts a terrific 13-6-3 record with a 2.59 GAA, a .921 save percentage and 4 shutouts in his first starting gig. The 28 year old journeyman is a big reason for the Colorado Avalanche’s turnaround from the Western Conference cellar in 2008-09 to first place in the Northwest Division as of today. Currently, Anderson ranks fourth amongst all players with 9.0 Goals Versus Threshold (GVT), though he’s lost some ground since ranking first two weeks ago.
Puck Prospectus readers already knew Anderson was good, even before his breakout season as Colorado’s starter, as he posted an impressive 10.1 GVT with the Panthers in 2008-09, and that was in only 27 games! As a result, Anderson’s 0.37 GVT per game was third to only Boston’s Tim Thomas and Florida starter Tomas Vokoun amongst goalies that played at least 20 games. After arriving in Colorado in the offseason, it didn’t take long for Anderson to unseat disappointing incumbent Peter Budaj (.902 save percentage and -9.0 GVT in 105 career games), who has subsequently seen action in only three games.
Since being a sub-replacement backup for the Blackhawks from 2002-03 to 2005-06, Craig Anderson has established himself as a premier goaltender – one who’s even being discussed as the possible starter for Team USA––over Buffalo’s Ryan Miller––for the upcoming 2010 Winter Olympics in Vancouver.
So where’s the complaint?
If you read our preseason projections for ESPN Insider, you probably came across the factoid that the otherwise excellent netminder was the worst active goaltender in the shootout going into this season. Exactly how bad had Anderson been? Well, in a league where stopping two thirds of shootout attempts is commonplace, allowing 12 goals on 20 career attempts for a .400 save percentage is astonishingly poor – small sample size be damned.
Fast forward to 2009-10. Watching the October 10th shootout between Chicago and Colorado––Anderson’s first taste of sudden death this season––I figured that the worst active goaltender would be a disaster. Sure enough, Jonathan Toews and Kris Versteeg scored with ease against a belly-flopping Anderson, both using nearly the same move; in boxing, Anderson would have been TKOed as “a defenseless foe” at this point. The shootout ends with Anderson going a worse-than-expected 0 for 3 if final shooter Patrick Kane just emulates Toews and Versteeg, committing Anderson to the ice, while roofing the puck on the stick side. But when Kane inexplicably doesn’t follow the blueprint, the shootout evolves into a nine round affair. Eventually, Anderson loses, but ironically ends up boosting his career shootout stats by stopping a pedestrian 5 of 9 attempts (.555), many against the soft underbelly of the Blackhawks’ shooters. Needless to say, sabermetric sirens went off in my head. How could a save against the lumbering Dustin Byfuglien (0 for 3 career) be equivalent to a save against the dynamic Patrick Kane (11 for 22 career)?
Clearly, to get a true sense of a goaltender’s (or shooter’s) shootout ability, the quality of their opponents needs to be taken into account. Using Craig Anderson as our Exhibit A, let’s first examine if he was really as bad as the .400 save percentage made him out to be before October 10th:
Craig Anderson - Career shooting percentage of opponents, 2005-06 to 2008-09
Shooter Result SA GA S%
Vincent Lecavalier Goal 40 12 30.0%
Brad Richards Goal 36 16 44.4%
Markus Naslund Goal 31 10 32.3%
Jarkko Ruutu Save 18 8 44.4%
Richard Park Save 10 3 30.0%
Martin Erat Goal 22 8 36.4%
Paul Kariya Goal 25 12 48.0%
Vyacheslav Kozlov Save 42 24 57.1%
Marian Hossa Save 33 12 36.4%
Richard Peverley Save 11 3 27.3%
Martin Erat Goal 22 8 36.4%
Markus Naslund Goal 31 10 32.3%
Nikolai Zherdev Goal 33 13 39.4%
Adam Hall Goal 1 1 100.0%
Evgeni Artyukhin Save 3 1 33.3%
Jussi Jokinen Goal 48 26 54.2%
Ryan Malone Goal 7 1 14.3%
Alexei Kovalev Missed 31 11 35.5%
Andrei Kostitsyn Missed 14 3 21.4%
Andrei Markov Goal 9 4 44.4%
Total 12 of 20 (60.0%) 467 186 39.9%
As alluded to above, a big factor affecting a goalie’s actual save percentage is how deep his shootouts go. If a goaltender is involved in only short shootouts, he faces the opposition’s top three shooters (assuming that their coach knows what he’s doing). These shooters typically average a 40% success rate as opposed to the 33% overall rate in shootouts*. Sure enough, before this season, Anderson faced four shooters only once, in a December 26, 2008 loss to the woeful Tampa Bay Lightning. Not surprisingly, over the 20 attempts, he only faced a sub-30% shooter three times (Ironically, allowing Ryan Malone’s sole career goal for the game winner in the aforementioned loss to Tampa). As the average shooting percentage of his opponents was 39.9%, we can deduce that Anderson was effectively better than his .400 save percentage showed. We should hope so.
*From the beginning of the 2005-06 season through November 23, 2009, 1553 goals have been scored on 4712 shootout attempts, or 33.0%. Annual rates have fluctuated between a high of 33.7% in 2008-09 and a low of 32.5% in 2007-08. To date in 2009-10, only 124 goals have been scored out of 399 attempts, or 31.1%, but this rate should be expected to climb up towards the historical rate of 33% over the course of the season.
By taking Anderson’s .400 actual career save percentage times the 39.9% average career shooting percentage of his opponents and dividing by the 33.0% average shooting percentage of all attempts from 2005-06 through 2009-10, we arrive at an effective career save percentage of .484 for Anderson, through the end of last season. If Anderson had instead faced a set of average 33% shooters, he would have been expected to register a .484 save percentage – an improvement, but still the worst career mark of all active goaltenders in the shootout.
Now let’s look at Anderson’s opposition in 2009-10, in 26 attempts that more than double our overall sample size for the Av’s netminder:
Craig Anderson - Career shooting percentage of opponents, 2009-10
Shooter Result S G S%
Jonathan Toews Goal 15 9 60.0%
Kris Versteeg Goal 4 2 50.0%
Patrick Kane Save 22 11 50.0%
Patrick Sharp Save 18 5 27.8%
Dave Bolland Save 4 0 0.0%
Dustin Byfuglie Save 3 0 0.0%
Troy Brouwer Save 2 0 0.0%
Tomas Kopecky Goal 2 1 50.0%
Andrew Ladd Goal 2 1 50.0%
Jason Williams Goal 14 4 28.6%
Henrik Zetterberg Save 31 11 35.5%
Ville Leino Save 1 0 0.0%
Mikko Koivu Goal 37 16 43.2%
Martin Havlat Missed 16 3 18.8%
Marek Zidlicky Save 13 4 30.8%
Patrick Kane Goal 22 11 50.0%
Kris Versteeg Save 4 2 50.0%
Patrick Sharp Save 18 5 27.8%
Tomas Kopecky Save 2 1 50.0%
Andrew Ladd Missed 2 1 50.0%
Cam Barker Save 2 0 0.0%
Troy Brouwer Save 2 0 0.0%
Dustin Byfuglien Missed 3 0 0.0%
Jonathan Toews Goal 15 9 60.0%
Patrick Kane Save 22 11 50.0%
Patrick Sharp Goal 18 5 27.8%
Total 9 of 26 (34.6%) 294 112 31.2%
This season has been marked by two marathon shootouts with Chicago (8 rounds and 9 rounds), which exaggerated Anderson’s relative success through defending against the likes of Dave Bolland, Cam Barker, Troy Brouwer and Dustin Byfuglien (0 for 11 combined). Predictably, Anderson’s opponents have been a below average group at 31.2% shooting percentage, a significant factor in his greatly improved .654 mark; a calculation of Anderson’s effective save percentage finds yields a .618 mark for 2009-10. Though greatly improved from his previous effective save percentage of .484, the two marks at least converge more than the actual save percentages.
Finally, we can take a look at where we would rate Anderson overall in career shootout ability:
Craig Anderson - Actual and effective shootout save percentages
Years SA GA Opponents' S% Actual Sv % Effective Sv %
Prior to 2009-10 20 12 39.9% 0.400 0.484
2009-10 26 9 31.2% 0.654 0.618
Career stats 46 21 34.9% 0.543 0.576
Taking into account all of Craig Anderson’s 46 career shootout attempts against, we find an effective career save percentage of .576; given slightly above average opposition averaging 34.9% career shooting percentage, Anderson is effectively a tad better than his actual career save percentage of .543. Without adjusting other netminders similarly, the .576 rate would still rank him in the bottom 10% of all goaltenders. And even if Anderson improved in ability to the .618 effective save percentage demonstrated this season, he’d still rank in the bottom third.
Yes, Craig Anderson has been great for the Colorado Avalanche this season, but when it comes to shootouts, we now have a pretty good idea of exactly how bad he has been.
Timo Seppa runs the statistical hockey site Ice Hockey Metrics.
Timo Seppa is an author of Hockey Prospectus.
You can contact Timo by clicking here or click here to see Timo's other articles.
| 
0 comments have been left for this article. 
Conn Smythe Watch: Lundqvist Leads