By Jack Schroeder
Around 4:00 am EST Saturday morning, skip John Shuster led Team USA to a 10-7 victory over Sweden in the men’s curling gold medal match in the 2018 Pyeongchang Winter Olympics. The result appeared near impossible just days earlier, as the Americans slumped to a 2-4 record in round robin play. Of the ten teams that qualify, only the top four qualify for medal rounds, meaning the United States needed to win all of their remaining matches (and probably get some help through tiebreakers). Although the American men had never beaten Canada at the Olympics, they did so twice on their route to the final. Importantly, three of the Americans’ six wins came without the initial hammer.
The team with hammer throws the last stone in each end (think: inning), which is a large advantage in curling. Susquehanna’s International Group’s Raise Your Game blog found that the team with hammer has scored in 90% of ends this year. During a match, hammer goes to the team that lost the last end. Figuring out who gets to start with hammer is a bit more complicated. After warming up, each team throws two stones, one released with clockwise rotation, the other counter-clockwise. Their distances from the button (center of the house [think: bullseye]) are added to make up the Last Stone Draw number. The team with the lowest LSD gets initial hammer, referred to on scoreboards as LSFE (Last Stone in First End). As a result, in contrast to a coin toss, the initial advantage of hammer is determined by skill. I sought out to determine how much of an advantage LSFE was in Olympic curling.
To answer this question, I made a historical database of men’s curling results at the Olympics dating back to the 1998 Nagano Olympics. I separated games by year to locate any potential trends and also calculated average points and margin of victory aside from winning percentage.
I found that LSFE teams win 63% of all matches. There are two main reasons behind this. First, being able to put points on the board early forces the other team to play from behind. Only 6% of teams with LSFE gave up points in their first end this year. Setting the tempo of play can have large ramifications, especially if the team with hammer can score more than one point. The second advantage comes from the fact that each Olympic match is ten ends, with ties being broken in an extra end. If the team with hammer wins a point every end, as is expected, LSFE means hammer in the extra end. That being said, many teams will intentionally blank the first end (ensure no stones are in the house when the end is over, letting them keep hammer) to put themselves in the position to have hammer in the tenth end. Since 1998, 32% of teams with LSFE blanked the first end. While that number rose to 46% in Sochi for the 2014 Winter Olympics, this year saw the lowest percentage of first ends blanked ever. It is too early to tell if this year was an aberration or a suggestion that strategy is changing over time.
One thing that is definitely changing over time is how often the team with LSFE wins their match. During the 2002 Salt Lake City Olympics, teams with LSFE only won 55% of their games. That number rose to 71% this year, reflecting a steady upward trend. Many commentators have expressed how high-level curling has become much more competitive in recent years. As Aileen Geving, an American curler, told The New York Times earlier this month, “In the past, if you were a good shot maker, you could be pretty good and still be successful. But the game has gotten so precise: The ice is better, and the players have gotten better. Any bit of edge is going to make a difference.” She believes the increased fitness of curlers is that difference. While this effect is visible in how curlers look–skip Shuster lost 35 pounds after being cut from the team after Sochi–the data suggests that LSFE is joining fitness and technique as determinants of success at the Olympic level.
We can isolate this advantage when looking at average points. As a whole, teams are scoring less than a stone more per match now than in 1998. Point average has been relatively volatile within the 5.5-6.5 range, however, and there does not seem to be a trend one way or the other since the field expanded to ten teams in 2002.
Still, when separating point average into LSFE and FSFE (First Stone in First End), the margin between the two has increased over time. What was half a stone’s difference in 1998 has quadrupled: the margin is now almost two whole stones.
A possible counterargument to the “evolution of the game” argument is that an expanded field has heightened the disparities between different teams and that LSFE only signals which team is better. I would contend that Pyeongchang largely uprooted this argument, with powerhouse Canada failing to medal. However, when analyzing initial hammer’s impact on a nationwide basis, all but two countries benefit from LSFE (who both suffer from playing very few Olympic matches, leading to sample size issues). While the impact of initial hammer is lower for better teams, the difference is still meaningful because initial hammer is binary: teams shoot first or second. Also, since the tournament is a round-robin, every team will play the other. If LSFE only signals high quality, there should be a large disparity between good and bad teams. Six out of the ten teams this year finished either 5-4 or 4-5, and the United States won gold with a 5-4 record, both of which imply that the curling ladder is not fully hierarchical.
In the future, it would be worth developing a database that did a better job of taking team strength into account when evaluating the impact of LSFE. While strength of schedule would be meaningless in a round-robin field, I would be curious to see how often weaker teams won with initial hammer. World rankings could potentially be used to get a better indication of a team’s strength, as using final records could dilute the sample. While mixed doubles does not currently have a large enough sample size to evaluate, it is worth considering whether LSFE has a large impact in women’s curling at the Olympic level.
Overall, LSFE appears to have a sizable impact on Olympic men’s curling. The impact of initial hammer has grown over the last few Olympic cycles as the sport has become more precise. Fitness has probably played a role in setting teams apart, but LSFE is important now more than ever. Team Shuster’s route to the gold medal this year was described by many as the “Miracurl on Ice.” Indeed, having to win three elimination matches without initial hammer–the penultimate round-robin match against Switzerland to go 4-4, the rematch against Canada in the semifinal, and the gold medal match against Sweden–should qualify as miraculous.