26: Women’s worlds review

2021.05.11

The Women’s Worlds are done and congratulations to the Swiss for winning it all. (I imagine Melanie Barbezat reading this and being like “Gee, thanks Ken. That means a lot.”)

The Swiss were clearly the best team over the 12 days of Worlds but they weren’t the best team on Sunday. I kind of like the concept of game ranks, which answers the question, how well did a team’s performance rank among all games played in an event?

I’ll use my homegrown points over expected for the rankings although using raw percentages will tell a similar story. Switzerland’s 4-2 win over RCF ranked 30th-best during the event. That’s out of 194 games played by all teams combined. So it was a solid performance relative to the field, but it was just the ninth-best game of Switzerland’s 15 contests. They gave the Federation some gifts and the Federation often refused to take them. RCF’s game ranked 64th for the week.

The best performance on Sunday was by the U.S. in securing a bronze medal against Sweden. Their game ranked 13th overall. It was the second-best game of the event by a team not from Switzerland, Sweden, or Russia. The best was also by the U.S. whose win over China held up as the 12th-best performance of the week. Maybe if the U.S. had played Switzerland on Sunday instead of Saturday, the result would have been different.

As it is, Team Tirinzoni moves to the top spot in my world ratings and will hold that place entering what will hopefully be a more normal 2021-22 season.

Now onto the more esoteric stuff from the Women’s Worlds.

Don’t complain to Marie Turmann about facing tough shots

Per my expected difficulty calculation, Estonia’s Marie Turmann had the three most difficult draws in the event:

Opponent     End  Shot  Exp pts.
Czech Rep. 2 16 1.69
Korea 3 13 1.82
Korea 3 15 1.92

Needless to say, she missed them all. Actually it’s not needless to say because the expected points algorithm is just doing its best with the limited information it has. And the shot against the Czech Republic was actually makeable. Turmann “just” needed a draw to full four-foot and there was a path to do so. Those draws against Korea, though…she really had nothing.

Don’t complain to Anna Kubeskova, either

With that in mind, it’s interesting to look at which players had the highest average degree of difficulty during the entire event because these things definitely don’t even out. Here’s the top-5 based on the average expected points on each player’s shots.

Player          Diff    POE
Kubeskova CZE   2.90   -0.06
Constantini ITA 2.91   -0.14
Turmann EST     2.92   -0.16
Han CHN         2.97   +0.13
D Jentsch GER   2.99   -0.02

In almost every event, this list is populated by fourths on teams that played from behind a lot. In the Worlds, due to the spread of talent, there’s also a component related to how good the first three (or two in Germany’s case) players were. Less-skilled players tend to not leave easy shots for their fourths.

Naturally, there’s a relationship between the quality of a fourth and the quality of the first three players, and so most players dealing with a high degree of difficulty tend to underperform the expectation, as we see on the above list. But there was one exception.

If Han Yu ever gets some help, watch out

It’s with that long-winded explanation that I wonder if you were impressed with China’s Han Yu as you should be. She faced tougher shots than the name-brand fourths and still did quite well. Here’s how she ranked among all fourths in points over expected (POE).

Player          POE
Paetz SUI      +.22
Hasselborg SWE +.16
Muirhead SCO   +.14
Han CHN        +.13
Kovaleva RCF   +.05

That’s impressive. Give Han some better results with her teammates’ first six rocks and this team will be a factor in the Olympics. Here were the positional ranks of her team:

1 Jiang -.04 (10th)
2 Dong  -.03 (9th)
3 Zhang -.01 (9th)
4 Han   +.13 (4th)

What’s interesting about Han’s week is that whether out of necessity or preference, no fourth threw more draws than her.

Player          Draw% Shots
Tirinzoni SUI 100.0 1
Han CHN 46.0 235
Dupont DEN 43.2 257
Kubeskova CZE 42.5 226
Constantini ITA 42.1 235
Yoshimura JPN 42.0 250

This list is also populated by teams playing from behind, so it was probably more necessity than preference. Still, Han often came up big in order to carry her team to a 6-7 record.

The Swiss played the best but also did the dumbest thing

Wait, what is Silvana Tirinzoni doing on that list? I didn’t see it reported anywhere – not that it was especially newsworthy – but it appears the Swiss had an incident in the eighth end against Sweden where Alina Paetz threw out of order.

This was in the midst of a nightmarish end where Switzerland would concede their only loss of the event after giving up four. Throwing out of order happens, of course, but it always amazes me how four players could forget how many rocks are left. Even more so when it involves the best team in the world. It occurred during the TV blackout, so it’s possible that it’s a bookkeeping error by the WCF, too. But in the official record, it’s real.

It was also the only time Switzerland surrendered more than two points in an end all week. Also, they gave up just seven steals in 15 games, each for a single point. Total dominance.

(Honorable mention on the blooper reel is Kerri Einarson and Co. peeling a rock with their fifth shot against the Czechs.)

That’s all, folks

That’s a wrap for this season. You’ll have to fend for yourselves on the World Mixed Doubles and U.S. Nationals. To close, let’s look back on all of my hypothetical wagers during the bubble and see how my ratings held up.

Scotties: Homan (+500, lost in final)
Brier: Bottcher (+500, won); Jacobs (+375, didn’t make playoffs)
Men’s Worlds: Scotland (+350, lost in final); Switzerland (+1000, finished 3rd); USA (+2000, lost in quarters)
Champions Cup: Mouat (+500, won); Kovaleva (+1000, lost in quaterfinals)
Players Championship: Gushue (+650, lost in final); De Cruz (+3300, didn’t make playoffs); Kovaleva (+1000, didn’t make playoffs)
Women’s Worlds: Kovaleva (+1000, lost in finals)

If you had put $1 on each bet, you would have spent $12 and won…$10. So there was no profit to be had, but we’ll do better next season with more games/data and more experience.

As for the offseason around here, there probably won’t be much posting. But I’ll be doing stuff behind the scenes to make the shot data and associated metrics more accessible. And I’m hoping to build a shot-based win probability model from this data as well.

Anyway, thanks to everyone who stopped by and read my work. I really enjoyed doing it. Let’s meet back here in a few months.

User comments
  1. Josh · 2021.05.11 · 11:56 am

    Thanks for doing this over the last several months. It’s been great to read analysis of curling like this that really can’t be found anywhere else!

    1. kenpom · 2021.05.11 · 5:28 pm

      Thanks, Josh!

  2. Mike · 2021.05.11 · 3:42 pm

    Ken, this is amazing. I don’t know how you found the time and energy to spin this up but I’m now an avid doubletakeout.com reader and maybe even curling fan despite having watched a total of maybe two curling matches in my life.

    1. kenpom · 2021.05.11 · 5:28 pm

      Thanks, Mike! That’s great to hear.

  3. Frank Mercogliano · 2021.05.11 · 4:06 pm

    You probably didn’t realize I’m VERY into curling (have a few local titles here!)

    1. kenpom · 2021.05.11 · 5:16 pm

      I definitely did not know this. It’s in the twitter bio and everything! I’m bringing my gear next time I’m in town.

  4. nuclearbdgr · 2021.05.11 · 7:09 pm

    Thanks Ken – it is great to have some extra context during the journey – your daily grid of who has played who was incredibly useful

    1. kenpom · 2021.05.14 · 12:18 pm

      Thanks, nb!

  5. Brice R. · 2021.05.12 · 7:36 pm

    Thanks for putting this together, Ken!

    This has been a blast to follow, and I’d definitely pick up a second subscription to support a full curling stats/analytics database.

    1. kenpom · 2021.05.14 · 12:19 pm

      Oh, wow. We are probably many years away from an audience big enough to support a subscription service, but someday…

    2. Brice R. · 2021.05.14 · 7:08 pm

      You’re probably right, sadly, but I figure it’s better to make that suggestion early 🙂

  6. Gerse · 2021.05.14 · 11:07 am

    This has been great, Ken – thanks for making it a thing!

    1. kenpom · 2021.05.14 · 12:20 pm

      Thanks for reading!

Leave a Reply to Gerse Cancel reply