Final Four at 2026 World Curling Mixed Doubles

Sweden and Italy won playoff qualification games to advance to Friday’s semifinal round at the World Mixed Doubles curling championship in Geneva.

Defending champions Italy – Stefania Constantini and Amos Mosaner – trailed 3-1 early against Japan’s Tori Koana and Go Aoki but scored the 7-6 win in an extra end.

It was the second straight extra end game for Italy following their win over Scotland to complete the pool matches.

Sweden’s Therese Westman and Robin Ahlberg scored four to start their match against Katie McMillan and Angus Bryce and added a deuce in the third end to lead 6-2, eventually winning by a 7-5 count.

The semifinals are set, with Italy taking on Australia (Tahli Gill and Dean Hewitt) and Sweden challenging Canada (Kadriana and Colton Lott).

Meanwhile, one question remains – how well did the pre-tournament model perform?

Below is a comparison between projected playoff probabilities and actual round robin outcomes.

Pool A – Predictions vs Results

Team                       Probability               Outcome

Australia (Gill/Hewitt) 77%                1

Japan (Koana/Aoki) 66%                  3

Sweden (Westman/Ahlberg) 58%   2

Estonia (Kaldvee/Lill) 43%                4

Norway (Kjærland/Brænden) 27%  5

China (Han/Yu) 22%                            6

Pool A followed expectations quite closely.

The top three teams all reached the playoffs, and the overall order remained largely consistent with the pre-event projections. The group featured well-established teams with stable performance levels, which tends to reduce volatility.

Estonia, projected as a borderline playoff team, ultimately finished just outside the top three. Their path remained realistic until the final stages, which aligns well with the model’s expectations.

Pool B – More Volatility, Bigger Misses

Team                      Probability                 Outcome

Korea (Kim/Jeong) 90%                     5

Italy (Constantini/Mosaner) 80%    2

Canada (Lott/Lott) 40%                     1

Scotland (McMillan/Bryce) 33%      3

Finland (Immonen/Sipilä) 14%          9

USA (Kawleski/Kauffman) 13%         10

Pool B was a different story.

While Italy, Canada, and Scotland all performed in line with expectations and reached the playoffs, the model clearly overestimated Korea, who missed out despite a strong ranking profile.

Below the top tier, results varied significantly. Several lower-ranked teams were able to challenge and occasionally defeat higher-ranked opponents, highlighting the inherent volatility of mixed doubles.

This is also where limited data and changing team dynamics play a larger role. Compared to Pool A, the field included more emerging or less consistent pairings, increasing unpredictability.

What Did We Learn?

The model performed well in structured environments, where team strength differences are clearer and more stable.

However, it struggled in:

Mixed doubles remains a format where variance is high and margins are thin. Even strong teams can fall outside the playoffs if a few key games go the wrong way.

Once the playoffs began, of course, projections become even less reliable.

Single-game elimination — or near-elimination — means that execution in the moment outweighs long-term indicators. A well-timed performance is often more important than a strong season.

Australia has shown consistency throughout the event and enters the playoffs in strong form. Canada, Italy and Scotland all looked like serious contenders for gold, while Japan and Sweden had the ability to upset any opponent on the right day.

At this stage, the model gives way to momentum.

And in mixed doubles, that can change everything.