1: Let’s talk curling

2020.07.07

Hi. My name is Ken Pomeroy. You may know me from my work in basketball. This is my site about curling. Thanks for stopping by.

This has been something I’ve wanted to do for a while. I watch a fair amount of curling. More than the average person, I’d say. And I find myself mostly lost on the power structure of the sport and things like when teams should adopt risky strategies.

There are questions for which I need to find answers. Who would be favored in a game between Tracy Fleury’s team and Eve Muirhead’s? Why do women blank more first ends than men? How much does a 10-end game increase the favorite’s chance of winning?

These questions may not seem important if you don’t know anything about curling. But one is not going to grow as an analyst if they don’t think about things outside their area of expertise. We are going to dive into a sport where the players are more loyal to improving their win probability than in any other sport where the most points determines the winner. There are things to be learned from that exercise that go beyond curling.

There have been some notable efforts in the numbers behind curling. I’ve learned a lot from Kevin Palmer’s Curl with Math blog which is the industry leader when it comes to the written word in curling analytics. And the ratings that will be used for much of the content here wouldn’t be possible without the data collection efforts of Gerry Geurts at Curling Zone. With a little digging you can find some academic papers on the topic, as well.

So it’s not that the curling analytics audience is underserved, necessarily. It’s obviously pretty small. Still, there is no handy reference for determining what Tracy Fleury’s chances are against Eve Muirhead. And if there was, maybe there would be more interest. I like surprises when I watch sports, but if I don’t know who is supposed to win then I can’t be surprised.

It should be noted that this work wouldn’t be happening without the coronavirus outbreak and America’s incompetent efforts to contain it. After the sudden cancellation of sports in March, I was suddenly left with a gaping and indefinite hole in my work schedule.

Throw in some civil unrest and my country is in the midst of one of its most depressing summers in modern history. So this is a project as much about my mental health as anything. I don’t encourage ignoring what’s going on. But at the same time, we all need a mental break and writing about an ice sport that might only be played in Europe and the Far East this season offers an escape to get as far away from reality as possible.

Given present trends, indoor sports in the U.S. may be limited for months, curling in particular. As a game played in a cold environment with a requirement of post-game social interaction, the sport combines the conditions of a meatpacking plant for the playing portion and a bar for the social portion. It’s the worst of everything in a world with a nasty virus that spreads among people before they show symptoms. Thus, it’s likely that recreational curling will be extremely limited in North America this season.

However, at the competitive level there’s some hope for action. Fans are not as integral to the curling experience as they are for the major team sports so that won’t hold events back. And in places like South Korea and Japan and China, where there’s been an effective national response to control the virus and there are a few elite teams, competitive curling figures to happen and will be streaming on YouTube at some point this season. Events might even happen in Europe and Canada, too – it looks like the Baden Masters is still a go in late August. It will not be a normal season, but there will be games somewhere.

I’m not yet sure what my voice will be here. But because the only person I can count on to read every post is me, my core audience is going to be me. That’s not someone who is a lifelong follower of the sport but it also isn’t someone who doesn’t know anything about it either. And I’ll be relating things to other sports that I have been interested in my whole life. So if you don’t know much about curling, you might not be lost.

But at some point, if you want to follow along, you’ll probably have to watch an actual curling match to have a rudimentary understanding of what the sport looks like. If you can watch only one game, I’d recommend the 2004 Brier final. But just catching 30 minutes of a random game on bonspiel will do the trick, too.

I do know that I won’t make this place all about technical jargon. I’d like to think my style is more about creative thinking and less about pure analytics. I’m not going to dump numbers and complicated formulas on you for the sake of doing so. In that regard I try to do the minimum to tell a coherent story.

That said, the team ratings on the front of this site will be the cornerstone for a lot of the discussion here. And before we go any further, I’ll have to explain how they work. So that will be the topic of next week’s post.

For now I’ll just say that plenty of interesting content can be derived from a decent predictive rating. That could be particularly useful for the upcoming season when there may not be much interaction among the best curlers from different regions. These ratings can still provide a robust approach to ranking the best teams in the world, even when their opponents are of wildly different quality.

I’m not sure how long this will last. I plan to write something weekly for at least a few weeks. If things are back to normal at that point we’ll see what happens. Or maybe this is the new normal. Who knows.

User comments
  1. Guy Beaudry · 2020.07.09 · 10:02 am

    Rankings look very interesting, particularly for the juniors. Will the list be sortable by Rank as well?

    1. kenpom · 2020.07.09 · 12:05 pm

      You can click on Rating and it will sort by rank. You can also click on Jr/flag/province to filter the rankings by those categories. Also, click on a team’s rating to see its chance of beating any of the other teams on the list. More to come!

  2. Shawn · 2020.07.09 · 11:27 am

    can you offer any more insight into how your rankings work? One immediate question, on the women’s side, you rank Tabitha Peterson and Nina Roth. As these teams are largely composed of the same players, how did this work out?

    1. kenpom · 2020.07.09 · 12:03 pm

      I’ll explain the ratings in detail in the next post which hopefully will clear up some things. For now, I am just rating the skips, which has its drawbacks. I plan to change that in the future, but in doing it this way it does show good Peterson’s team was at the end of the season. People knew they were good, but they were *really* good. The Peterson/Roth situation will be topic of its own post in a couple weeks.

  3. shawn · 2020.07.09 · 1:30 pm

    I suppose the answer will come in the next post you refer to as, but how do you rate the skips without taking into account the quality of the team?

    Somewhat related, when you refer to the quality of opponent, are you referring to the quality of opposing team, or opposing skip?

    On a side note, happy to see you involved in this. Love to see the BB ideology you use, think it is the gold standard as it is such logical based upon 100 possessions. Curious to learn how your thinking works on curling. Especially as you are more of a fan than someone who actually plays the game at a high level.

    1. kenpom · 2020.07.09 · 2:16 pm

      I’m just using the skip as the name of the team. Similar to how the WCT world rankings just list Jason Gunnlaugson to refer to his team. It’s just that I don’t recognize the games where Roth was gone as Team Roth like the WCT does. That is the most obvious case, but there are some other ones like Carruthers skipping for McEwen in an event last season. Those games get credited to a Team Carruthers which ultimately didn’t play enough games to get listed on my ratings.

  4. Shawn · 2020.07.10 · 6:59 am

    Your next post will probably answer this, but I have never been called a patient person.

    In basketball, you have an adjusted tempo, that takes into account pace of play. Do you have something similar in your curling stats? Maybe in curling it would have to be a conservative or aggressive factor? As an example, against a conservative skip, you cannot afford to miss more than a shot or 2, or you will lose 3 to 1, but against an aggressive skip, you might miss a few and give up big ends, but also get big ends back, so you end up winning 10 to 7.

    1. kenpom · 2020.07.10 · 10:42 am

      Yeah, you’ll have to wait until the next post. But it’s pretty simple and just about who you play and who you beat.

  5. Jon · 2020.07.10 · 10:26 am

    I’m only familiar with curling from the Olympics but I really enjoyed it and am happy to learn more. I take it from the ratings that competitive curling outside of the Olympics is primarily an individual sport. Are there team competitions as well, and are you considering team ratings? I would find that more compelling to follow honestly. Thanks.

    1. kenpom · 2020.07.10 · 10:45 am

      Curling is entirely a team sport and these are team ratings. It’s just that the teams don’t have names and are traditionally identified by the player that plays skip.

  6. GregJP · 2020.07.10 · 11:14 am

    Hi Ken. I’m a Canadian who has watched curling for over 40 years and am also a stats geek and retired Math teacher. I just wanted to tell you that what you are doing here is awesome. I can’t wait for your thoughts going forward.

  7. GregJP · 2020.07.10 · 11:17 am

    One thing that I’d be interested in seeing you talk about is how big an upset Shuster’s Olympic gold medal was compared to upsets in other sports.

    1. kenpom · 2020.07.10 · 4:25 pm

      Sure! That’s something we can do with a system like this.

  8. Shannon · 2020.07.10 · 3:50 pm

    Awesome! I can’t wait to follow along 🙂
    Thanks for sharing with all of us.

    1. kenpom · 2020.07.10 · 4:25 pm

      Thanks for stopping by!

  9. Ed · 2020.07.11 · 9:47 am

    I asked this on Twitter, but: is there any way those of us who played limited competitive schedules could get the unlisted data on our own teams? My team (Scimia) only played 9 tour games on the season, but I’d be curious to know what our “provisional” rating was/is, just for fun.

    1. kenpom · 2020.07.11 · 11:37 am

      Not sure how I missed that tweet. If you saw my reply to Sussman, your rating is almost identical. If I ranked every team that played at least one game, you’d be around 750th.

    2. Ed · 2020.07.11 · 9:38 pm

      Thanks Ken! Great work, looking forward to seeing more discussion in the weeks to come.

Join the discussion!