Dan Ellis vs. UFA Back-ups

Early in the season, many fans thought the Hurricanes found a solution to their long-standing problem of acquiring a suitable back-up to Cam Ward in Dan Ellis. Originally signed to a professional tryout contract with the Charlotte Checkers, Ellis played tremendously during the NHL lockout and earned the back-up job in training camp. His performance in the first few months of the season were very good, as he got to start a handful of games and did a fine job in relief of Ward in a few games. As a back-up, Ellis did just about everything you could ask from a goaltender, but he was presented with a much bigger workload after Ward suffered a knee injury that would keep him out for the rest of the year. Ellis would have to go from being a back-up goalie to a 1 A/B type role and start more games than he was used to. His performance in this role can be divided into two parts. 

Prior to Ward's injury, Ellis had a save percentage of .925 over six games and he ended the year with a save percentage of only .906, which is a little less than what you would expect from a typical back-up. Ellis definitely struggled in the latter half of the year, but he also went through some injury troubles of his own. Before the Hurricanes 4-1 loss to the New Jersey Devils on March 21, Ellis posted a save percentage of .912 over six games after Ward's injury. Ellis himself also sustained an injury this game after cutting his leg open with his own skate blade and this would keep him out of action for two weeks. After he returned from the injury, he posted a save percentage of .887 in his final seven starts of the season. Prior to the that, his save percentage was .918 over 12 games.

It is my opinion that Ellis was not fully healthy after returning from injury because before that occurred, Ellis' performance was adequate and more than suitable for a back-up goaltender. If the Hurricanes are going to be searching for a #2 goalie this summer, going back to Ellis would not be a bad option. Justin Peters might be under contract, but he has not proven himself to be a capable NHL goaltender and with $6+ mil. invested in Ward, the Hurricanes don't have the cap space to spend friviously on this position. Since the Hurricanes haven't been able to develop a back-up goalie (yet), they will have to turn to free agency to find a back-up netminder and retaining Ellis seems like an easy thing to do for next year. How does he compare to other available back-ups, though?

I've already taken a look at how Ellis compares to other back-ups in terms of yearly save percentage and he ranked favorably, but we're going to use a different metric today courtesy of Hockey Prospectus. Save percentage is usually a reliable stat for judging goalies over a period of time, but it has it's issues when you use it over a small sample size, which is the case when you look at back-up goalies. Most back-up goaltenders play about 20% of their team's games and a lot of them are weeks apart, so small sample size issues can make going by save percentage along misleading. This is why I like looking at how often each back-up goaltender gave his team a chance to win, which can be found by looking at how many "Quality Starts" he had.

This stat was derived by Rob Vollman by Hockey Prospectus and it judges goalies based on how often they gave their team a chance of winning. The methods are explained well in this article by Lighthouse Hockey, but the jist of it is that teams are expected to win 70% of the time when their goalie stops at least 91.2% of the shots he faces. Therefore, any goalie who posts a save percentage of .912 or above is awarded a "quality start." Quality starts are also awarded to goalies who post a save percentage between .885 and .911 depending on how many shots they faced and how many goals they surrendered, as teams are expected to win at least 50% of the time when their goalie posts a save percentage above .885 in a game.

In addition to that, any goaltender who fails to stop 85% of the shots he faces earns what is known as a "blow-up," giving his team a very low chance of winning. These types of performances are usually the most back-breaking since it usually involves one team being forced to play from behind. Carolina fans know all about this since they were blown out in many games in the latter half of the season and a good chunk of them were due to sub-par goaltending. Although, not many of these games were Ellis' fault and he actually gave the Hurricanes a chance to win in over half of his starts.

Goalie GP Quality Blow-Up QS% BU%
Ward 17 10 3 58.8% 17.6%
Ellis 19 11 4 57.9% 21.1%
Peters 19 7 4 36.8% 21.1%

Ellis gave the Hurricanes a solid chance of winning in 11 out of his 19 starts and only put them in a hole four times and two of those came after he was injured. I'm not sure how many games Ellis would be expected to start in an 82-game season, but if he can give the Hurricanes a chance to win in 60% of his starts, then that is mighty fine for a back-up. He actually gave the Canes a chance to win in roughly the same percentage of starts as Ward, which is why I am sick of hearing about how his injury "doomed" the Hurricanes.

If anything, Ellis' injury is what caused a lot of their goaltending woes. I mentioned earlier that Ellis rushed himself back from a pretty bad looking injury and he posted only two quality starts after returning from that. To make matters worse, Peters gave the Hurricanes a chance to win in less than 50% of his starts and that helped put the Hurricanes out of all contention. Ellis' performance after the injury is concerning, but he still has a better recent track record compared to other options in free agency.

Goalie GP Quality Blow-Up QS% BU%
Emery 21 14 1 66.7% 4.8%
Ellis 19 11 4 57.9% 21.1%
Labarbera 15 7 0 46.7% 0.0%
Garon 18 8 5 44.4% 27.8%
Theodore 15 5 5 33.3% 33.3%
Mason 11 3 5 27.3% 45.5%

Based on how often each goaltender gave their team a chance to win, Ellis ranks favorably compared to other unrestricted free agents. The only one who did a better job was Ray Emery, who might be available depending on what Chicago does with him. He also has a sketchy history with injuries, so he isn't the most stable option out there even if he is potentially the best goalie on the market. Jason Labarbera is also not a bad choice for a back-up goaltender and he probably has the best recent track record of any UFA goalie here. This season was the first time in the last five years that Labarbera was not quality in half of his starts and even then, he didn't have one terrible performance in 15 games, which is typically what you want out of a back-up. Labarbera is experienced as a #2 goaltender, so he may not be a bad replacement for Ellis if the Canes choose to not keep him. 

I'm not sure what the team's final decision will be when it comes to finding a back-up, but re-signing Ellis for another year might be their best option. When healthy, he was an ideal back-up for the Hurricanes and he has proven over his career that he is capable of starting at least 20-25 games, which is all Carolina will ask of him when Ward is healthy. With limited cap space and no one within the organization to turn to, we could see Ellis return as Carolina's back-up.

Stats courtesy of Hockey Prospectus and Hockey Reference.