Martin Brodeur spoke to the media recently about the state of goaltending. He said that goalies today were being “babied” and not playing back-to-back. This post explains why that is not a good idea as well as pushing back on the quote.
Since the New Jersey Devils played in Toronto on Tuesday, the Toronto-focused part of the hockey Media focused on the Devils. Pierre LeBrun had an interesting chat with Devils VP and legendary goaltender, Martin Brodeur, about the state of goaltending in the NHL for TSN. Brodeur definitely showed his age somewhat with the following comments about goalie usage. Per TSN:
“It’s the system of 1A and 1B that’s going to create that,” Brodeur continued. “Because it’s always going to be, ‘We can’t play back-to-back (games).’ You’re so scared the No. 1 gets hurt. And so you baby that guy for the longest time.
“So these guys are playing 55 games. They should be playing 65 to 70 games like we did.
“I think we baby our goalies. I see it. I’m part of it. It’s like, my goalie coach will say, ‘He’s played five games the past eight days …’ I’m like, ‘So?’”
Brodeur is very much a legend and I intend no disrespect. However, this take is pretty bad if only for how it does not acknowledge the changes in his position. Given Brodeur’s role in the front office, I think he really needs to do so.
The part that stuck out to me the most was about back-to-back games. I can understand where Brodeur is coming from. It has become a rarity of sorts. And for good reason. The risk is not just that the goaltender may get hurt, but that they may perform worse and hurt the team’s chances at winning that second game of a back-to-back. Let us dive into matter.
The Current State of Goalies Playing Back-to-Back
NHL.com has goalie stats by days rest. In this case, 0 days of rest would be of our interest. This would mean the goaltender played the very next night after getting some action. I have pulled the numbers from the site prior to Wednesday’s games on March 27.
Right away, we can see that only 27 different goaltenders have had to do that this season. This is 27 goaltenders out of the 94 who have taken to the crease even for one second in 2023-24. Out of those 27 goalies, only six have played with no days of rest between games more than one time. For a total of 34 instances out of the hundreds of games played this season so far. That right there shows how rare it has been.
We have to acknowledge that this includes cases where goalies are coming in to replace someone, whether due to injury or poor performance. In the context of Brodeur’s quote about back-to-backs, he is referring to starting a goaltender in both games. That is an important distinction. The move away from playing goalies on consecutive nights was referring to starts. To make that distinction, I went through those 34 instances of a goalie playing on both nights. Out of those 34, only 12 involved a goalie starting both games. The other 22 involved a goalie coming in relief in one of the two games.
How well did it go for those 12 instances in 2023-24? Let us look at the all situations data from NHL.com and Natural Stat Trick to find out. This is a New Jersey Devils blog so let us look at the Devils first. Who, for some reason, owned three of these instances. They lead the league in this kind of usage too.
Akira Schmid
Akira Schmid owns the distinction of being the only goalie in the 2023-24 season to start back-to-back games on two separate occasions. He also acts as a great cautionary reason why teams should not do this.
Game 1: November 30, at Philadelphia – 3 goals allowed on 47 shots for a save percentage (Sv%) of 93.62% with 4.2 xGA. Devils won 4-3.
Game 2: December 1, vs. San Jose – 5 goals allowed on 17 shots for a 70.59% Sv% with 1.52 xGA. Devils lost 3-6.
This first set really irks me. Schmid faced a flood of rubber in that Philly game, did rather well, and then suffered for it by being thrown onto the ice the very next night. The Sharks preyed on Schmid. Even if the team was and is terrible and, at the time, winless on the road, Schmid and the guys in front of him were not fully prepared to play and they suffered one of their worst losses of the whole season because of it. I may not be an expert, but this is Exhibit A as to why a team should not start a goalie right after facing a one of work the night before.
Game 1: December 16, at Columbus – 3 goals allowed on 26 shots for an 88.46% Sv% on 2.55 xGA. Devils won 6-3.
Game 2: December 17, vs. Anaheim – 4 goals allowed on 27 shots for an 85.19% Sv% on 2.54 xGA. Devils lost 1-5.
While Schmid did not face a ton of work against Columbus, it was work. Not the best work but manageable nonetheless. The Devils indeed won that game. Then, like their San Jose game earlier in the month, the squad as a whole laid an egg to a vastly inferior opponent. In retrospect, starting Schmid on both nights was a mistake. A costly one if you believe the rather cold take that losing games to teams like Anaheim are a reason why the Devils are not in the playoffs.
Nico Daws
Not learning anything from Schmid, the Devils decided to put Nico Daws in back-to-back games in February. This actually worked out well despite the risk as Daws was hot at the time. The hotness would last for two more games after this.
Game 1: February 12, vs. Seattle – 1 goal allowed on 28 shots for a 96.43% Sv% on 2.36 xGA. Devils won 3-1.
Game 2: February 13, at Nashville – 2 goals allowed on 32 shots for a 93.75% Sv% on 3.34 xGA. Devils won 4-2.
This could be used as a reason why the risk is just that: a risk. A goalie could still do well and a team could still win. Which happened in this back-to-back set. Daws’ numbers were still a bit worse in that second start in Nashville, although he faced more shots by both volume (32) and danger (xGA) in that one. I wonder whether this is what Brodeur was talking about when he questions why a goalie cannot play in this situation. I would point out that this was the only time this happened with Daws this season.
That covers 3 out of 12. The results of the other 9 were mixed.
Joonas Korpisalo
Game 1: March 23, at New Jersey – 2 goals allowed on 20 shots for a 90% Sv% on 3.01 xGA. Ottawa won 5-2.
Game 2: March 24, vs. Edmonton – 3 goals allowed on 36 shots for a 91.67% Sv% on 3.96 xGA.
Korpisalo managed to have a better game in his second straight start a somewhat rare occurrence. Although it is worth noting that he just had 20 shots to contend with in the afternoon before the Edmonton game.
Scott Wedgewood
Game 1: January 20, at New Jersey – 2 goals allowed on 31 shots for a 93.55% Sv% on 3.23 xGA. Dallas won 6-2.
Game 2: January 21, at NY Islanders – 3 goals allowed on 28 shots for a 89.29% Sv% on 2.49 xGA. Dallas lost in overtime 3-2.
The ex-Devil goalie stood tall in the Brylin Debacle. Dallas found it fit to give him the game in Long Island on the next day. That did not go as well. Not terribly, but a downgrade even with fewer shots and less dangerous shots as a whole. The Stars left with one point instead of two.
Pytor Kochetkov
Game 1: December 6, at Edmonton – 3 goals allowed on 6 shots for a 50% Sv% on 0.5 xGA. Pulled after 5:36 of the game. Carolina lost 6-1.
Game 2: December 7, at Calgary – 3 goals allowed on 27 shots for a 88.89% Sv% on 3.34 xGA. Carolina lost 3-2.
Pytor Kochetkov and Carolina’s goaltending earlier this season were a total mess. This back-to-back start was more of hoping Kochetkov would be better after a miserable start in Edmonton. It was not like they had a lot of faith in Antti Raanta either. Kochetkov was by all accounts better in the second game since he stopped more than 50% of the shots he faced. He cleared a bar set underground for how low it was. Still, Kochetkov was not exactly brilliant either and the Canes dropped both games in Alberta.
Logan Thompson
Game 1: December 27, at Anaheim – 5 goals allowed on 26 shots for a 80.77% Sv% on 2.68 xGA. Las Vegas lost 5-2.
Game 2: December 28, at Los Angeles – 2 goals allowed on 34 shots for a 94.12% Sv% on 2.8 xGA. Las Vegas won 3-2.
Las Vegas absolutely did not trust Jiri Patera to get the game after Thompson and Las Vegas were rocked by a four-goal first period in Anaheim. They went with Thompson and hoped he would be better than his awful first game. And he was. Thompson and the G-Knights provided a rare case of the second game both being a victory and an improved performance after a bad start.
Ukko-Pekka Luukkonen
Game 1: March 2, vs. Las Vegas – 2 goals allowed on 24 shots for a 91.67% Sv% on 2.33 xGA. Buffalo won 7-2
Game 2: March 3, vs. Winnipeg – 3 goals allowed on 33 shots for a 90.91% Sv% on 2.32 xGA. Buffalo lost 5-2.
Buffalo absolutely creamed Las Vegas in the first game and Luukkonen was perfectly fine. The second game – boosted by two empty net goals – was not a bad start for the goalie against the Jets. But it was another downgrade in a losing effort by his team compared with the first game. The risk was recognized and took place, even with a small degree.
Charlie Lindgren
Game 1: January 13, vs. NY Rangers – 2 goals allowed on 27 shots for a 92.59% Sv% on 2.12 xGA. Washington won 3-2
Game 2: January 14, at NY Rangers – 2 goals allowed on 31 shots for a 93.55% Sv% on 3.42 xGA. Washington lost 2-1
Charlie Lindgren, arguably Washington’s MVP this season, started a home-and-home against the then and now current division leaders. He more than did his job in both games. In fact, it is arguable that he did even better in the second game in light of more shots against and more danger presented by those shots. Washington lost that game but scoring just one goal will often lead to that. Mark this is an example of a back-to-back start working out for the team like Thompson, Korpisalo, and a few more goalies coming up on this list. The first is next.
Ilya Sorokin
Game 1: January 15, at Minnesota – 3 goals allowed on 32 shots for a 90.63% Sv% on 2.42 xGA. Pulled after 40 minutes of the game. NY Islanders lost 5-0
Game 2: January 16, at Winnipeg – 3 goals allowed on 43 shots for a 93.02% Sv% on 4.54 xGA. NY Islanders lost 4-2
Ilya Sorokin was more or less pulled in the first game after the second period to give his team a spark. Not only did it not work, Semyon Varlamov conceded two more goals for a bad loss in Minnesota. Sorokin would go on to have a better performance in the second game – which still led to a losing result. The issue of allowing 43 shots stands out more to me than the 3 allowed.
Connor Hellebuyck
Game 1: March 23, at NY Islanders – 5 goals allowed on 26 shots for a 80.77% Sv% on 2.48 xGA. Pulled after 29:11 of the game. Winnipeg lost 6-3
Game 2: March 24, at Washington – 3 goals allowed on 19 shots for a 84.21% Sv% on 2.24 xGA. Winnipeg lost 3-0.
The Jets and Hellebuyck were crushed in half of the game in Long Island back on the 23rd. The fifth goal allowed just before the halfway mark got the big money goalie pulled. He would have a better night in Washington. Not a good one, mind you. Allowing 3 goals on 19 shots still reeks of something, even if it was not as foul as the first game in this set. This is only technically an improvement.
John Gibson
Game 1: January 15, at Florida – 4 goals allowed on 33 shots for a 87.88% Sv% on 3.44 xGA. Anaheim won 5-4 in overtime.
Game 2: January 16, at Washington – 1 goal allowed on 27 shots for a 96.3% Sv% on 2.57 xGA. Anaheim lost 2-0.
Gibson amazingly went from being beaten in one game in a winning effort thanks to the guys in front of him to having an awesome game in a losing effort. Hockey is not fair. But Gibson rebounded well with such a short turnaround from the previous game. He did his job well. He can join Thompson, Korpisalo, Sorokin, Lindgren and technically Kochetkov and Hellebuyck’s whose second starts in a back-to-back ended up being better than the first start for them.
Martin Jones
Game 1: January 2, at Los Angeles – 0 goals allowed on 31 shots for a 100% Sv% on 3.32 xGA. Toronto won 3-0
Game 2: January 3, at Anaheim – 1 goal allowed on 28 shots for a 96.43% Sv% on 2.63 xGA. Toronto won 2-1 in overtime
Jones has the best back-to-back set of starts of the season in the NHL. He was perfect in the first game. He was near-perfect in a 1-1 game that had to be decided in overtime. Of course, giving up 1 goal out of 28 shots – which is great! – is a downgrade from a shutout.
Conclusions
Out of the 12 instances, about seven saw the goaltender post a better save percentage in the second game compared to the first. However, those seven included three goalies who were pulled in that first game including a Kochetkov performance that saw him beaten three times on six shots. If you take out the goalies who did not finish their game, then you are left with four times out of nine where the goalie did better in the second game than the first one.
Not that it necessarily led to a win either way. In the 12 instances of goalies going back-to-back with starts, the combined record was 5-7-1 after a combined first game result of 9-3-0. While a record is a team stat, goaltenders can and are chosen based on whether the team just won with them. Just look at Vitek Vanecek’s 29 starts and 32 appearances with New Jersey this season in spite of an overall save percentage of 89% (which was often below it for those games) for an example. Those 17 wins with him in the crease justified way more minutes than his miserable save percentage would.
The larger lesson is not that goaltenders should never start both games in a back-to-back. It can work out. Even if the second game was a downgrade, it did work out for Nico Daws and Martin Jones in the one time their respective teams did that. It also worked out well, performance-wise, for Lindgren, Luukkonen, and Korpisalo regardless of the result. It can work. It just does not often work. With more goaltenders capable of performing in the NHL, there is not much of a need to lean on one goalie to start back-to-back.
Again, this is about 12 cases out of hundreds in this season. If it was a good idea, then you would see teams do it more often. Like they did in Brodeur’s day that apparently had less babying involved.
Back in Brodeur’s Day, Goalies Did Play With No Rest More Often and It Wasn’t Always Good Idea Then Either
Let us go back to NHL.com. Thanks to their data history, we can look at goaltenders by days of rest even as far back as Martin Brodeur’s hey day. Let us look at a season from his prime years: the 1999-2000 season. An excellent season for the goaltender as well as the New Jersey Devils. In that season, Brodeur played and started in 72 games. Near as I can tell from his game log, he was only pulled in one of them too. Per NHL.com, Brodeur started in eleven back-to-backs in 1999-2000. He was clearly not held back then.
It also did not go so well for the legend. When Brodeur played with zero rest between games, he posted an overall save percentage of 88.2%. A big decrease from a 90.8% on one day of rest between games and much bigger compared with two days of rest when Brodeur put up a 92.1%. Brodeur and the Devils kept on winning even with lesser save percentages so it was fine. Plus, an 88.2% save percentage was better than the 87.6% that backup Chris Terreri had in 1999-2000 over his 12 appearances.
Brodeur was not wrong that goalies were not babied back then. Those 11 games where Brodeur played on zero rest was not even the most in the entire NHL that season. (Aside: Or the all-time leader as five all-time legends from the original six era have him beat.) Latvian legend Arturs Irbe led with 14 such instances for Carolina. Whereas Marc-Andre Fluery is this season’s leader in games played on zero rest with 3 – all three involved him coming in relief for one of the games – there were 33 goalies who have done it at least 3 times back in 1999-2000. A list that included big name goalies like Patrick Roy, Tomas Vokoun, Dominik Hasek, Ed Belfour, ESPN Breaking Newsman Kevin Weekes, and Mike Richter along with Brodeur. Those 33 included goalies that did play in tandems back then like Martin Biron & Hasek and Mike Dunham & Tomas Vokoun for example. Brodeur was right that it was done differently back then.
With the benefit of hindsight, we can point out that it often was a bad idea. Brodeur’s own save percentage with zero days rest was worse compared with having one or more days of rest. This applied to a lot of those goalies in that season as well. Irbe posted a 90.6% with no days of rest; a drop of 0.4% from having one day of rest. Richter was a 90.7% goalie with one day of rest; he dropped to an 89% goalie with no rest. Fred Brathwaite went from 90.7% on one day of rest with a steeper drop than Richter with an 88.5% with no rest. Biron went from a 90.5% save percentage on one day of rest to a miserable 87.2% save percentage. Tommy Salo had a similarly big and bad difference going from a 91.8% save percentage on one day of rest to a 88.9% save percentage on zero days of rest. Guy Herbert went from a 91% on one day of rest to a 89.6% on no days of rest. When you face hundreds of pucks, a difference of 0.1% can mean quite a bit. Drops of a full percentage or more is huge. Mind you, these are all goalies who played at least eight games on no days of rest. You can see some bigger drops down that list like with Roman Turek, Olie Kolzig, and Hasek.
Were there exceptions? Sure. Somehow, Weekes was somehow 2% better at stopping pucks with no rest compared with one day of rest. Jocelyn Thibault, Stephane Fiset, and Ed Belfour saw big gains in save percentages when playing on no rest compared with one day of rest. However, those were exceptions. A lot of goalies did play a lot worse with no days of rest. This was just one season in time too. It is not the only one either. And the saving grace for some of these improvements or not-steep falls in percentages is that it includes goalies coming in relief. Which has less to do about goaltender choice and more to do with just seeing through games often already decided.
For the most part, in history, goalies playing on consecutive nights did not go so well for their own numbers and that would impact the teams. That is why teams have went away from that kind of usage to where we stand now. It is a team’s responsibility to use their players as best as possible for the team to win games and put the player in a place to succeed. We have learned from the past that starting goaltenders back-to-back is a real risk for a team to take and it does not always pay off. Therefore, the league has corrected itself and now it is a rarity today.
Additionally
One of the big parts of goaltending back in Brodeur’s day was the butterfly. While Brodeur himself did not often go into a full one, a lot of goalies did. Inspired by Patrick Roy’s usage to great success in Montreal and Colorado plus the influence of goaltending coach Francois Allaire, a lot of goalies in the 1990s followed the form to a ‘T.’ It was also a costly measure as it led to a lot of hip injuries as well. In these more recent times, a new style has emerged: RVH or Reverse Vertical Horizontal. Popularized by Jonathan Quick for covering pucks, the RVH has proven to be effective as well as put a lot of wear and tear on the body. Hip injuries continue to be a concern along with groins, knees, and ankles. And with goalies growing up playing in these ways, these injuries can and do happen well before they even hit the pros – which adds further risk to the player.
While efforts are ongoing to determine how to best mitigate those risks in how they tend goal, both forms and other reasons have ultimately led to teams managing their goaltenders a lot differently now than they did 20 years ago. Brodeur’s lament about the lack of goalies who can play in 65 to 70 games is understandable in that he played in an era where that did happen. And, for some like Brodeur, it was expected and appreciated. But with teams now having at least two goaltenders of quality on top of risking further injuries due to overwork, it makes little sense to just run goaltenders over and over in this season.
Brodeur is right that it has gone away. Brodeur himself last played over 65 games with 77 appearances in the 2009-10 season. He did not play more than 59 games in the following five seasons he was active in. Which makes sense as he was definitely getting up their in age and his effectiveness was not what it once was. Back to the point, there have been just 55 individual seasons of goalies playing 65+ games since Brodeur last did it. Only four of them happened within the last five years: Carey Price’s 66-game season for Montreal in 2018-19, Devan Dubnyk’s 67-game season for Minnesota in 2018-19, Connor Hellebuyck’s 66-game season for Winnipeg in 2021-22, and Juuse Saros’ 67-game season for Nashville in 2021-22. No one did it at all last season as Saros and Hellebuyck fell one game shy of 65. Unless Saros plays in eight of Nashville’s last ten games or Alexandar Georgiev plays in nine of Colorado’s last ten games, no one will do it this season either.
Could they play more games? Maybe. I guess. But why would Nashville or Colorado would want to do that? I do not think Nashville or Winnipeg would be substantially better off if Saros and Hellebuyck were approaching a 70-game season. Going back to the back-to-back section, I do not think it is a coincidence that Saros has not played on zero days of rest this season and Hellebuyck has only done it once. If the team trusts Kevin Lankinen and Laurent Brossoit to get the job done here and there – and they do – then that is the smarter play.
Regarding Brodeur’s comments stretches of games played by a goalie: I think that is absolutely worth exploring with the 2023-24 Devils as goalies have had multiple streaks of appearances. Did they come at a cost? I need to think about how to approach it before I can say for sure it did or not. I am inclined to think so given the state of the 2023-24 Devils goaltending.
Your Take
While I understand where Brodeur is coming from, he is pretty much stuck in the past about it. Which makes sense because he would use his own experiences to guide his thinking. That is natural, I think. However, I would hope that he understands why the reality of goaltender usage is the way it is. The wear and tear the position takes on a goalie plus the talent available as well as the results of past usage has all contributed to why things the way they are. No one is doing any “babying.” They are making choices based on information.
Starting goaltenders back to back remains a risky decision and often not a successful one. It does not happen much in this league and the Devils have had it work (as in not blow up in their face) one out of three times. For the few times it happened elsewhere in the league, the results for the goaltender or their team were mixed. I do not think the solution to that is that it is not done enough. It definitely was not the solution to do it with Schmid twice in the same month. Further, I do not think it should be the view of someone in management that a goaltender is being treated too lightly or being babied because they are not playing 8+ games with zero rest between days. Given how poor the Devils’ goaltending has been this season, I wonder if I have to question whether the front office has played a role in that beyond Tom Fitzgerald failing to get a goaltender when the People Who Matter have increasingly wanted one since November.
However, this is my take. Now I want to know yours. Why should a team intentionally start a goaltender in back-to-back games? What would make it work? Will we see an eventual sea change in how goalies get used? Please leave your answers and other related thoughts in the comments. Thank you for reading.