5: The transactional blank I

2020.08.06

Between the dismissal of Lisa Weagle, the taking advantage of residency rules, and the burned-rock incident at the Olympics, Rachel Homan has her critics. I can’t say I’m a Homan fan, either, but I do have a certain appreciation for a team that walks onto the ice and consistently crushes its opponent with all the enthusiasm of a trip to the grocery store.

I must say that after digging through some data, I also admire how Homan has a particular philosophy about playing. Specifically, she believes the first-end blank is very valuable for the team with last rock. She believes that more than any other team in the world.

First-end with hammer, percentage of blanks
(top-ten teams, last two seasons)

1  Homan       53.3
2  Hasselborg  34.5
3  Fleury      28.1
4  McCarville  27.3
5  Kovaleva    24.8
6  Tirinzoni   20.0 
7  Muirhead    18.3 
8  Jones       17.2
9  Einarson    16.8
10 Fujisawa    15.7

I’m only listing the top ten teams here on account of laziness, but even if we expanded to the top 25, or the top 100, or every team in the known universe, Homan would be at the top of the list in first-end blanks. If you’re tuning into a Homan game and she has hammer, you might as well have some alternative programming queued up for the first 15 minutes.

If Rachel is consistent then she should be equally obsessive about her opponent blanking with hammer in the first. And she is. In fact if any reader can produce video evidence of Homan calling for a first shot in the rings, I’ll reward you with a personal mention. When Homan throws first, you get Lisa Weagle throwing a tight guard that on rare occasions slips in to the rings.

First-end without hammer, percentage of blanks
(top-ten teams, last two seasons)

1  McCarville  44.4
2  Muirhead    34.3
3  Hasselborg  27.7
4  Fleury      27.6
5  Kovaleva    25.6 
6  Tirinzoni   22.4
7  Homan       18.9
8  Jones       18.6
9  Einarson    17.8
10 Fujisawa    15.5

This is mix of 8- and 10-end games but in the 10-end Canadian provincials and the Scotties, first-end blank percentage is through the roof. In 8-end games only, Homan’s opponents blank just 12.8% of the time, easily the lowest among top-ten teams.

That’s not the lowest in the world, though, because it requires some skill by both teams to pull off the blank. So there are less-skilled teams with some pretty low blank rates, especially among juniors. But among world-class team she’s definitely on the low end. Anna Sidorova, the 12th-ranked team in my ratings, only blanks 9% of the time without hammer in the first end of an 8-end game.

It’s not a stretch to say the most boring first end in curling is Homan with hammer and the most exciting is Homan without hammer. In the latter case, she’s going to put up a guard and as the end progresses, play with fire more often that other top teams, men’s or women’s.

The commitment to this strategy is admirable but I’m not going to run for the vacant presidency of the Homan Fan Club unless the approach helps the team win. One way to evaluate that would be to look at all of Homan’s games over the past two seasons. This is how she’s done with hammer along with her record in each case:

Score   W-L    Pct
 All   90-17  84.1
 -2     0-0    --
 -1     7-1   87.5
  0    48-9   84.2
  1    15-7   68.1
  2+   20-0  100.0

Homan has won 84.1% of the time with hammer and when she blanks the first end she has won 84.2% of the time. So at first glance, all those blanks aren’t doing much. If you’re playing Homan, the first-end blank seems to be an acceptable outcome.

The benefits are perhaps in the cases where Homan doesn’t blank. The defensive strategy prevents many steals. She’s only been stolen in 8 of the 107 games and all of those were single points. And while the first end is usually boring when Rachel has hammer, the rest of the game is boring if she happens to score two or more. She’s gone 20-0 in those situations (though I’m sure some of those games were still worth watching).

Anyway, the sample size is a bit small and includes games against lesser foes, too. It might be more telling to see what happened in all games between elite (top 15) teams the last two seasons.

Score   W-L     Pct
 All  239-189  55.9
 -2+    6-11   35.3 (-20.6)
 -1    27-39   41.0 (-14.9)
  0    70-50   58.3 ( +2.4)
  1    71-67   51.4 ( -4.5)
  2+   65-22   74.7 (+18.8)

I’ve added the difference in win probability from the overall average because we can use that information to evaluate the first-end-with-hammer performance of any team. The next step is straightforward. Take a team’s first-end scores over the past two seasons (with hammer and against top 15 teams only) and add up the win probability gained or lost based on the chart above. Voila:

             Net Wins
             per 100G    GP
1  Muirhead    +3.1      28
2  Kovaleva    +1.6      33
3  Homan       +1.5      39
4  Einarson    +1.3      41
5  Fleury      -0.5      54
6  Tirinzoni   -1.2      35
7  Hasselborg  -1.3      43
8  Jones       -3.5      31
9  Fujisawa    -4.6      24
10 McCarville  -5.3       4

Adding up all the bits of win probability based on first-end scores reveals that Eve Muirhead has gained 3.1 wins per 100 games from her first-end performance against top 15 teams over the past two seasons. Homan’s performance looks good here, also, and the only reason it’s not better is that she gets forced to one 20.6% of the the time. That’s the best figure among the elite teams, but barely so. It’s not as low as one might expect considering that fewer than half of her ends had a score in them.

It’s also worth noting that Jennifer Jones’ aggressive first-end approach hasn’t fared well. That’s not proof that the strategy is poor, though. These numbers reflect the results of strategy plus execution. Einarson seems to be similarly aggressive with better results. And also, the strategy may be more useful against weaker opponents than elite teams.

Still, most top teams on the men’s and women’s tours will gladly keep the end simple if their opponent throws the first shot in the rings, accepting a small bump in their chance of winning. The question, then, is why does the team without hammer so often throw the first shot in the rings? Homan’s massive blank percentage could not have happened without some cooperation from her opponents. While you’ll have to spend a lot of time finding a case where Homan calls for the first shot in the rings, her opponents do it all the time.

Throwing the first shot in the rings just leads one down the path of conceding 2-3% of win probability, maybe more against a team like Homan. Opponents are making it slightly harder on themselves to win while making Homan as comfortable as possible since she is more willing to play this way than any other team in the world.

Why are teams without hammer, especially on the women’s side, so willing to offer the blank to their opponent? We’ll talk about that next week in part two of our series on the transactional blank.

User comments
  1. Chris Allen · 2020.08.08 · 5:03 am

    I’ll take a stab at the “Why do so many teams let Homan blank the first” question. If I were the underdog (as most of Homan’s opponents are), I would think that effectively shortening the game by one end might benefit me. Kind of like how, say, Campbell would have a (slightly) better chance of beating Duke in a 35-minute game than in a 40-minute game. That begs the question…obviously a first end blank benefits the hammer team if both teams are equal, but if the team without first end hammer is a large enough underdog, is there ever a point where a first end blank benefits the non-hammer underdog more than the hammer favorite? I am sure you will get the bottom of that.

    1. kenpom · 2020.08.08 · 3:05 pm

      This might be the reason, but I wonder if it’s a good one. I think the effect of shortening the game in curling is different from other sports and I’ll probably do a separate piece on that at some point. The benefit to shortening the game in basketball is increasing variance, but in curling some shots have very little variance.

      I think that explains why elite curlers are much better when tied with hammer with 1 end left than with 8 ends left. They get to play peels and open hits which they hardly ever miss. Even tick shots have little variance these days. To increase variance you need to make your opponent play high variance shots. You need to take a lead on Homan to make her play those shots and blanking the first end gives you one less chance to get hammer back so you can do that.

Join the discussion!