Starting Goalie Career Stats

Updated: January 28, 2013 at 11:29 am by Ryan Campbell

As many of you have noticed and commented on, we have stopped posting goaltenders career statistics versus their opponent when we confirm them. There are two very simple explanation for this:

1) Fantasy managers should not be using this information to decide who to sit/start.

2) If you do want this information, it can easily be found on each goalies Yahoo! player page. Stats vs. Next Opponent are listed on the main page.

Let’s tackle explanation number one. There are three main reasons why past results have little to no correlation to future performance:

1) The only thing teams have in common over a several year span is the logo on the front of the jersey. Players change, coaches change, systems change.

To use an extreme example, Martin Brodeur has career record of 46-30-9 against the Philadelphia Flyers, posting a 2.44 GAA and .906 SV%. These numbers are well below his career averages of 2.23 and .913. However, 90% of that data in this set is irrelevant for the reason mentioned above. His first game against the Flyers was on March 29, 1992, a game which the Flyers won 5-4. As this game occurred over 20 years ago, you are probably not surprised to learn that Brodeur is the only player from this game still in the NHL. Here are the game rosters:

New Jersey Devils: Tommy Albelin, Laurie Boschman, Tom Chorske, Zdeno Ciger, Pat Conacher, Ken Daneyko, Bruce Driver, Bill Guerin, Alexie Kasatonov, Claude Lemieux, Troy Mallette, Randy McKay, Jarrod Skalde, Scott Stevens, Kevin Todd, Claude Vilgrain, Eric Weinrich, Valeri Zelepukin, Martin Brodeur, Craig Billington (Brodeur got hooked), Tom McVie (Coach)

Philadelphia Flyers: Keith Acton, Brian Benning, Claude Boivin, Rod Brind’Amour, Dave Brown, Terry Carkner, Al Conroy, Kevin Dineen, Mark Freer, Garry Galley, Mark Howe, Kerry Huffman, Dan Kordic, Andrei Lomakin, Mark Pederson, Dan Quinn, Mark Recchi, Mike Ricci, Dominic Roussel, Bill Dineen (Coach)

The statistics from this game have absolutely no bearing on how many goals Claude Giroux will score on Brodeur in their upcoming game on February 15, or how well Anton Volchenkov will play in front of him. In fact, when this game was played, Sean Couturier and Adam Larsson weren’t even born yet.

2) Goalie’s individual circumstances and talent level change

I think this goes without saying, but 1992 Brodeur, 1998 Brodeur, and 2013 Brodeur have different talent levels. The  stats he put up in his prime are not terribly relevant to today’s Brodeur, who doesn’t quite have the same physical talents as when he was young.

Goalies also play on teams of varying talent levels from year to year. An excellent examples of this is Roberto Luongo. For his career, Luongo has a 11-11-2 record, 2.77 GAA, and .913 SV% against the Atlanta Thrashers/Winnipeg Jets. However, the the majority of these games occurred when he was a member of the Southeast Division, and played for a perennially weak Florida Panthers squad. Luongo is now part of a very strong Vancouver Canucks team that has been among the NHL elite five of the last six seasons. Luongo won’t play against the Jets this year unless they meet in the Stanley Cup Finals, but if they did, I would wager that he has a much better than 50/50 shot of coming out as the winning goalie, and would likely have better than a 2.77 GAA.

3) The game changes year to year, and stats are not normalized across eras

In 1992-93, only 4 goalies posted a SV% above .900, with Curtis Joseph’s .911 mark leading the way. League average was 0.885. Last season, THIRTY goalies posted a better mark than CuJo, and league average was .914. Again, this is an extreme example and Brodeur is the only goalie remaining from 1992-93, but we don’t have to go back very far to understand how quickly the statistical landscape of the NHL can change. In 2003-04, NHL games averaged 5.136 goals, and the average SV% was a record .922. Post-lockout, 2005-06 numbers jumped to 6.050 and .906 respectively. We are now approaching pre-lockout levels, with 2011-12 SV% checking in at .921.

Goaltenders who began their careers in 1999 are going to have a lot better career statistics than goalies who broke into the league in 2005 because they had an opportunity to accumulate stats when the goal-scoring environment in the NHL was depressed. It’s all relative to league average. A quick exercise:

Which of these goaltenders is better?

Goalie A: 1,029 GP, 2.54 GAA, .910 SV%

Goalie B: 472 GP, 2.28 GAA, .920 SV%

Kind of a trick question as A is Patrick Roy, one of the best goaltenders ever, and B is Henrik Lundqvist, arguably the best post-lockout goaltender. Lundqvist has much better numbers because Roy happened to begin his career in an environment when NHL games averaged over a goal and a half more than when Lundqvist did. Despite posting weaker raw numbers, Roy lead the league in GAA four times and SV% three times. Lundqvist has never lead the league in either category, and can only dream of matching Roy at this stage in his career.

I have clearly used some extreme examples to illustrate my point, especially in bullets 1 and 2, using a wide range of dates that are only applicable to goalies who have been around for an extended period of time. However, circumstances can just as easily change in 1-2 years. From 2001-2006, the Pittsburgh Penguins finished dead last in the Atlantic division each season, averaging 62.5 points per season. In 2006-07, top picks Marc-Andre Fleury, Evgeni Malkin and Sidney Crosby had matured into stars and they jumped from 58 points to 105. Their goal differential swung by 103 goals. When evaluating goaltender matchups that season, even data from the previous season was next to useless.

Career stats versus an opponent make for fun story lines, but story lines don’t win fantasy championships. By all means feel free to look them up, and if you want to use them in your decision making process, that is your decision. But be critical – weigh recent performance more heavily, evaluate coaching and personnel changes. We will do our best to provide a nugget or two of information in our confirmations, but career stats will not be that nugget.

Henrik Lundqvist, Martin Brodeur, Patrick Roy