Mark Johnson once pulled his goaltender when tied in overtime. (Dan Sanger)
Most sports have a strategy or two that is risky and would usually not be attempted were it not for dire circumstances. A team needs points immediately or it is going down in defeat. In football, it’s the Hail Mary pass, going for it on fourth down, or the onside kick. Basketball has the desperation heave or fouling to force the other team to make free throws. Baseball is odd because there isn’t a clock that is running out, so the search for end-of-game offense usually involves sending up a pinch hitter.
Of the three, hockey’s desperate measures are more similar to those in baseball than basketball or football. A baseball team substitutes for the batter that is due up; a hockey team substitutes a regular skater for its goaltender.
Do these devil-may-care strategies work? Obviously, they must work some of the time, or they wouldn’t be attempted. Taking football in particular, I have very vivid memories from my youth of being depressed for weeks in the wake of a game involving an opponent first converting on fourth and long, and then completing a Hail Mary. I’m sure there are some Green Bay fans who will carry similar scars induced by an onside kick.
As for women’s hockey fans in the United States and Canada, yes, we can recall a game where an extra-attacker goal played a pivotal part, even if it didn’t involve college hockey or take place on this continent.
How often is the strategy successful in women’s Division-I play? According to the Women’s National Statistics Database for this season’s games played through Jan. 20, there have been 88 empty-net goals scored by the 34 teams that compete with the aim of reaching the National Collegiate Tournament, compared to 22 extra-attacker goals. All of those goals are not necessarily on an equal footing; some may also be power-play goals, while others occur when the teams would have been skating four-on-four had one not pulled its goaltender.
Does that same ratio of four empty-net goals for every one extra-attacker goal hold for prior seasons? Sadly, no. Using the same statistics source, last year there were 92 goals scored into an empty net but a whopping 50 goals scored with an extra attacker on the ice. For the 2012-13 season, the database shows 93 empty-net goals and 52 with an extra attacker. Part of the discrepancy appears to be that the extra-attacker goal sum also includes some that we would more accurately describe as delayed-penalty goals.
For example, Clarkson’s Shelby Nisbet scored at 19:40 of the first period in the national championship game. The box score denotes that goal with an “EA.” While it’s true that Erica Howe had been off for 10 seconds and a sixth skater was on the ice, that was due to a delayed penalty on Minnesota, which was assessed once Nisbet’s goal was scored. While the two goals are the same in many respects, in the case of the delayed penalty, the attacking team is not in jeopardy of being scored upon, so it belongs in a different bucket in terms of a metric when we try to determine the risk versus reward of pulling the goalie for another skater.
It’s not entirely accurate to say that an attacking team is not at risk of yielding a goal during a delayed penalty. The most recent example I’ve seen of that came on Oct. 22, 2011 in Duluth. The referee spotted an interference infraction by Wisconsin’s Saige Pacholok. As Minnesota-Duluth was in possession of the puck, play continued and Bulldogs goaltender Jennifer Harss made her way to the bench. A sixth UMD skater came on. After 20 seconds, a Bulldog attempted to make a diagonal pass back to one of her points. The pass failed to click, and the puck bounced off the side boards at neutral ice and slid into the empty net in the UMD zone. Because Wisconsin had not touched the puck in the sequence, no whistle saved the Bulldogs. In an ironic twist, the official scorer credited the unassisted goal for the Badgers to Pacholok.
At the time, Minnesota-Duluth coach Shannon Miller said, “I’ve never even seen that before in the 25 years I’ve been coaching.”
Getting back to what the ratio should be for the number of empty-net goals compared to extra-attacker goals, from personal experience, I’d say the four-to-one ratio for the current season would be more likely than the less than two-to-one ratios for each of the prior two years. The absence of a goaltender figures to make a team far more vulnerable than a sixth skater will make its offense prolific. The accuracy of the number of true extra-attacker goals can be improved slightly with a couple of filters. If only third-period goals and those scored by a team that is trailing are included, then the number of extra-attacker goals drops from 52 to 46 for two years ago, and from 50 to 39 last year. There may still be a handful of delayed-penalty goals in the count. More could be squeezed out by looking only at goals scored at the end of the third period, but as we’ll discuss later, that may also discard some bona fide extra-attacker goals.
In any case, there are certain situations where every coach in the country is going to elect to pull the goaltender in favor of an extra skater. Down a goal with 30 seconds remaining in the game, both teams at equal strength, faceoff in the offensive zone, that goalie is coming off if she isn’t out of the game already.
Beyond that, coaches have to forge a strategy out of a combination of what they believe the percentages favor and a hunch in the current game. If the coach waits too long, the risk is that no more goals are scored before time expires, and the status quo favors the team holding the lead. If the goaltender is pulled too soon, the coach invites an empty-net goal from the opponent, and the hole gets deeper.
To find which teams are most likely to gamble, we can look at the number of minutes teams play with an empty net. In that respect, coach Chris Wells and St. Lawrence traditionally rank near the top.
“I’m of the mindset that what’s the difference between 3-1 and 7-1,” Wells said. “We lose to Clarkson, 5-0, two of the goals were empty nets. We lose to Dartmouth, 5-1, in Rochester, and two of those are empty nets, and Brown scored an empty-netter on us. Once it gets to a certain point, I’m not trying to make a mockery of the game or the score of the game, but to me 5-3, 6-3, 7-3 is no difference, and that’s the mindset of our team. We’ve come to enjoy playing with the goalie out and the desperation of it.”
He gets no argument from current Saints captain Amanda Boulier regarding the strategy.
“I can remember some games where there will be seven or eight minutes left and ‘Wellsy’ will pull her, and no one even thinks anything of it,” Boulier said. “It’s like, ‘Alright, let’s do this; here we go.’ It keeps things interesting, and I wouldn’t want it any other way.”
There’s usually method behind such madness.
“My first year [coaching SLU], a playoff game against Clarkson, we were down two [goals] with five minutes to go,” Well said. “We pulled the goalie; we scored two goals in the last five minutes, and then we won in overtime on the Friday night game, and then we won on Saturday. Then most recently at Quinnipiac two years ago in the playoffs, we won the first night in overtime, 1-0. Then the second night, we’re down 2-0 with 29 seconds to go in the game with a faceoff in their zone, and we scored. We scored with .7 seconds left to tie it. Then we eventually lost in the third overtime. We’ve had a fair amount of success. I think at one point, we were close to being even with the goalie out and goals against.”
To have that kind of success, a team has to execute.
“We get some good puck pressure, and I think the desperation helps out, and the ability to get on pucks,” Wells said. “It’s hard to stick it in the empty net; it really is.”
Minnesota found that out in its NCAA quarterfinal versus North Dakota in 2012. Trailing 5-0 with 5:53 elapsed in the third period, UND pulled goaltender Jorid Dagfinrud while on a five-on-three power play. North Dakota scored with those six skaters playing versus three to cut its deficit to 5-1, and coach Brian Idalski proceeded to yo-yo Dagfinrud in and out of the net over the remainder of the game, partly because it had to kill a couple penalties of its own. UND played without a goalie for over seven minutes, and the Gophers failed to take advantage. The final score was 5-1, so there was no damage done, but it serves to reinforce Wells’ point.
The scoreboard may have something to do with that. A week later with only a 2-1 lead, the Gophers scored just four seconds after Amanda Mazzotta left the ice to cement a semifinal win over Cornell.
Some goaltender pulls are more strategic than others. In another NCAA quarterfinal in 2012, Boston University trailed Cornell 7-4 with 11 minutes remaining. Coach Brian Durocher pulled Kerrin Sperry while on a power play, and the Terriers scored 12 seconds later. They added two more power-play goals to tie the game at 7-7 with 1:57 left in regulation without having to pull Sperry again. A couple hours later, Lauriane Rougeau scored for Cornell at 19:50 of the third overtime to make the BU comeback for naught.
Surprisingly, that’s how some of those dramatic comebacks end, such as the one Wells described where St. Lawrence fell in the third OT to Quinnipiac. Perhaps that is because after three overtimes, nobody can remember which team fought back from a deficit.
That’s not always the case. The most dramatic comeback I remember seeing involving extra-attacker goals came in the 2005 WCHA Championship. Wisconsin trailed 2-0 when Mark Johnson pulled Meghan Horras with 1:12 remaining. Carla MacLeod scored 22 seconds later to pull her team within 2-1. Horras returned, the Badgers got the puck deep, and she left again. It took only 14 seconds for MacLeod to tie the game, 26 seconds after her first goal. In the 24 seconds that remained before intermission, Wisconsin was assessed two penalties. When the teams came back for the first overtime, Krissy Wendell scored the winning goal on a five-on-three power play before anyone could learn whether the Badgers had gained momentum.
Wisconsin versus Minnesota games tend to create strange endings. In 2009, they went to overtime in their final head-to-head, regular-season meeting. Trailing the Gophers by two points in the WCHA standings, Johnson knew that winning a shootout and gaining one point likely wouldn’t be enough to gain the league crown, so he pulled Jessie Vetter in a tie game with 31 seconds left in overtime.
“Knowing that they needed the win, I figured that they would do that,” Minnesota coach Brad Frost said. “But yeah, it was different.”
In the teams’ most recent game this season, Wisconsin had a slim two-point lead in the standings, but the Gophers had two games in hand. Even if the Badgers won a shootout, Minnesota could take over first place by winning one of those two games, so UW needed to win outright. With 45 seconds left in overtime, Minnesota was assessed a penalty, so Wisconsin had an offensive-zone draw with both teams one skater short. This time, Johnson didn’t pull his goalie for a skater.
“I lost a defenseman early in the second period, and we just came off a good minute and 20 seconds of a penalty kill,” Johnson said. “I didn’t have enough fresh bodies to find the five kids that might be able to do that. So the thought was there, but I didn’t have the personnel to execute.”
While that would have been essentially the same as a five-on-four power play, six skaters versus five is something that teams use far less often. Maybe not all teams. During Wells’ tenure at St. Lawrence, the Saints have always played more than 25 minutes per season with an empty net. Three times, that total has been more than 40 minutes, including 48 minutes last year.
To be sure, some of that time results from delayed penalties. For an estimate, Minnesota never trailed at the end of any game during its perfect season, so it didn’t have any late-game goalie pulls, but the Gophers still played with an empty net for 4:32 that year.
Wells admits that it probably isn’t a good thing to always be near the leaders in empty-net minutes, because it is a sign his team is behind. But if SLU is going to opt for an extra-attacker that often, should they practice that along with other special teams?
“A couple of years ago we did have a setup that we worked at for a couple of years because we had the same personnel,” Wells said. “We haven’t done that yet this year, but we’ll probably get to that point to practice that. It’s fun.”