Coward’s curling


One of my favorite discoveries from shot data is that when teams tried a draw (as opposed to a guard) on the first shot of the game in the 2021 men’s worlds, they went 4-32 (11.1%). When teams tried to throw a guard on the first shot, they went 29-32 (47.5%). It’s pretty neat when the data indicates that a more exciting strategy is also a more successful strategy.

The general perception among experienced curlers is there’s value to throwing the first shot in and offering the opponent a first-end blank. You bank some time and get a feel for the ice. But over and over, in big events with dozens of games, one prediction you can safely make is that teams that throw in on their first shot will do worse than teams that try to play real curling from the get-go.

The same pattern appeared in the 2021 women’s worlds: Teams that tried to throw in went 14-35 (28.6%), and teams that tried to guard went 24-24 (50%). I could go on with the dozens of events for which I have acquired data but to put it simply: If you know nothing about the teams involved, just watch the first shot of the game. It’s a more useful predictor of victory than who has hammer!

In fairness, part (most?) of the reason for this is that the worst teams in the event are the ones typically throwing a draw on their first shot. Obviously, that single decision on how to start the game does not have an impact on whether you make shots in future ends. But it’s a signal. A signal that you’re a coward.

Even if you don’t believe it’s cowardly, the fact that the worst teams in an event are the ones using the strategy might tell you it’s a strategy to avoid. It’s also worth noting that in the two slams last year, where there really weren’t any mismatches, this effect persisted. Teams that tried to throw in went 6-23 (20.7%) and teams that tried to guard went 34-67 (33.7%).

Maybe that difference will shrink as we get more data, but it’s also noteworthy that in the slams, fewer teams threw in on their first shot. Part of that is playing 8 ends in the slams as opposed to 10 ends in the worlds, but I’d like to think part of it is the knowledge by elite teams that being tied without hammer requires some aggression, whether it’s the last end or the first.

It makes some sense that massive underdogs would want to throw in on their first shot of the game. A blank end is a win when there’s a possibility of giving up three or four by playing real curling. It’s a cowardly approach, but sometimes you just want to play a few ends before you get your ass kicked.

However, for teams with aspirations of greatness, it doesn’t seem to make much sense. To that end, credit to Brad Gushue. I have 49 games in my database where he started without hammer and in every one he tried to throw a guard on the first shot of the first end.

But not every great team plays this way. Having watched a bunch of first shots for one of the first posts on this site last season, it didn’t surprise me that Krista McCarville has been the most likely to play cowardly curling, guarding on just 2 of 24 first shots in three Scotties appearances for which I have data. McCarville is far and away the most extreme Canadian skip in this respect.

Being one of the best teams in the world, I’m sure McCarville has her reasons. To her credit, she’s gone 13-9 in the 22 games she’s opened with a draw. That’s the fourth-best winning percentage among the 50 skips that have at least ten games without hammer. So I doubt we’ll see a change in her approach.

Last week in curling: The men’s event in Oakville was almost pure chalk, with eight of the top nine teams in the field making the quarterfinals. (The only miss was #16 Korey Dropkin missing out and #19 Jason Gunnlaugson making it.) Then the top four teams in the field made the semis…and the top two made the finals…and the best team won it.

Anyway, we appreciated the coverage on and in particular the commentary of the affable John Cullen. But we’ll admit we got a bit tired of hearing how Brad Jacobs was the number one team in the world. Because in this house, Bruce Mouat was, is, and — after crushing the toughest field in curling so far without taking a loss — continues to be the number one team in curling, possibly for many weeks to come.

South Korea’s EunJung Kim won the Saville Shoot-Out, the marquee event on the women’s side, and in doing so moved from ninth to sixth in the world. I had high hopes for Kim in last season’s worlds as her track record before being forced out of the game in 2018 was that of a top-ten team. After a dismal start her team managed to finish seventh and just miss out on a free pass to the Olympics. However, in winning last week against a field that included Homan, Einarson, Fleury, and Jones, she clearly establishes South Korea as the best team headed to the Olympic Qualification Event in December.

Kim is back in action against a similar field this week in Sherwood Park and while she was able to avoid top-ranked Rachel Homan after Homan’s quarterfinal loss at Saville, event organizers have sadistically put Kim and Homan in the same pool this week. In fact, Homan and Kim will face each other in the first draw on Friday afternoon.

Finally, my favorite moment of the week was this boss move by Kim. She simply turns her back as her opponent delivers a stone. In my mind, Kim recognized that the only way this game could be fair was for her to not know what her opponent was doing. One could not have blamed her for thinking so.

User comments
  1. Doug Wilson · 2021.09.16 · 7:23 am

    Ken – A friend just sent me your site and i really like reading your thoughts. Quite, well, thought provoking.

    Re the stats about throwing guards or drawing in, could it be that the best teams want to junk it up, make the game complex and put the weaker teams under pressure right away? I mean, if you think you can outplay them, why not put lots of rocks in play and force them to play harder shots.

    Conversely, the weaker team wants an open, simple game with fewer rocks on play. Krista McCarville, who is awfully good but plays a` very limited part-time local schedule, might want to play a simple game when she’s at the Scotties against full-time opponents.

    Anyway, my point is that weaker teams keep it simple by drawing on, stronger teams junk it up by throwing guards. Stronger teams also win more because they’re, well, stronger. So i think there’s correlation but not causation to the guard/draw throwing. Fwiw.

    Doug Wilson

    1. kenpom · 2021.09.17 · 1:52 am

      Thanks for checking out the site, Doug. I love your work.

      I agree that worse teams are generally the ones employing this strategy. The question of whether it’s the best approach even for weaker teams is a fair question, though. Even in an open end, a more skilled opponent is rarely going to be forced and practically always will avoid a steal. An open game keeps the scoreboard close, but it doesn’t necessarily give the worse team a better chance for a win. It will take some further examination to pin down whether that’s the case.

Join the discussion!