Now Available! EBook from CWM

Now Available! EBook from CWM
Order Now from Amazon

You can also get an epub copy


Tuesday, February 28, 2012

Scotties Recap and Brier Preview

Scotties Recap

Congratulations to the Heather Nedohin rink from Edmonton.  A fantastic effort at the Canadian Women’s Curling Championship, aka The Scotties.  They beat both Manitoba and BC twice on their way to victory.  Despite playing well, they started 1-3 and nearly every game became a must win. 

The Saturday night Semi-final game which started only two hours after the Alberta win against Quebec ended with a touch of drama.  Tied with the hammer 5-5 in an extra end against Jennifer Jones from Manitoba, it looked as if Heather’s final draw slid a fraction deep. From the TSN camera, it appeared the Manitoba stone was closer to the pin.  The measurement took place and Alberta had the point by what appeared to be a few millimeters. The lesson, don’t ever trust those overhead cameras

Heather wins her second Scotties, the first as a skip.  Since her win as third with Cathy King (then Borst) in 1998, a few changes have taken place.  She married Dave Nedohin, a swarthy Winnipeg transplant who throws the stone almost as well as she does. If only she’d hyphenated, Godberson-Nedohin would look great on the back of a Team Canada jacket, no?  Add in motherhood to the mix, raising two girls while your husband wins a few of his own major events, and she may have begun to wonder if it would ever happen again.  Well done Heather.  And now, let’s look at a couple of strange decisions.

In the Semi-final versus Jennifer Jones from Manitoba, Alberta is leading 5-4 without hammer starting the 9th end.  Common practice is to play aggressive, towards the centre of the house, in an attempt to force your opponent to one and get hammer coming home.  You might even steal and, likely at worse you might surrender two and have hammer coming home one down.  Of course, you could also give up three.   

Heather chose to put her first rock in the rings, tempting Manitoba to hit and play out for a blank end.  What do I think of this?  In the men’s game, this is a horrible decision.  Top men’s teams win 75-80% when tied with hammer.  The chance to go from 60% to 80% is worth the risk of a possible drop to 40% if you surrender a deuce.  Also, a three is less common occurrence.  

In women’s, a team with hammer tied in the final end wins only 70% of the time.  However, one up without hammer is still 60%.  If Heather chooses to be aggressive in the 9th, she is attempting to gain only a 10% winning percentage.  If she could steal it jumps to 85%.  The risk, however, is a drop to 40% if Jones scores a deuce or 15% if they manage to score three.   

Looking at historic stats Jones scores threes about 10%, deuces about 25% and gives up steals about 15% of the time.  These are suspect as the data size is not very large and the variance of their competition is likely substantial (ie.  they likely score more points against weaker teams).  If we use them for an analysis, and assume no blank if she’s aggressive, Heather is facing this:

W (blank) = 60%

W (aggressive) = (.1)(.15)+(.25)(.4)+(.15)(.85)+(.5)(.7) = 59%

So, it looks almost equal.  What appeared to be a strange decision is actually fairly close.  Let’s adjust for Alberta being a stronger team, reduce Jones’ chances of a three or deuce and increase Heather’s chance at a steal:

W (aggressive) = (.07)(.15)+(.20)(.4)+(.18)(.85)+(.55)(.7) = 63%

The number still only goes to 63%!

I’ll admit I was dumbfounded by Heather’s decision and thought it was the wrong decision.  The numbers show if wrong, it’s only barley and one could argue that the reduction in variability, especially given the importance of this game, is worth a few percentage points in some rough math analysis.  

Well done Heather and congratulations on not only winning the Scotties, but stumping the CWM guy!

Now why you put two rocks in the rings in the 10th end is a question I can’t answer with math.  I cannot understand the strategy but expect you have a good explanation.  Maybe it was not the original intent but an adjustment after the first rock came only to the top twelve?  Since TSN always misses the first 2-3 shots of every end, viewers may never know.

Other notes:

I need to ask the CCA if coaches are required to sit in a specific spot (I expect that is the case).  During the odd ends, including the critical 9th and extra end, coaches walk the length of the ice, eating into a 90 second timeout.  This is comparable to a basketball coach coming down from the 15th row and missing half the team’s timeout.

2012 Saskatoon Brier Preview

Kevin Koe of Alberta versus Glenn Howard of Ontario, the rematch.  Most people will be expecting this to be the final game to determine the 2012 Brier champ on Sunday, March 11th.  Gushue and Fowler are the next closest followers with Cotter and Jacobs also possible contenders.  Anyone else would be a surprise and perhaps this will be the year to have one.

Howard has a dominant record over the years, but most recent battles have been very close games, including a tight battle during the 2010 Olympic Trials and a great Thursday night game at the 2010 Brier, written about here  The final game of that 2010 Brier was one of the best ever, with Kevin drawing to the button to defeat Howard 6-5 in an extra end.  Koe's Alberta rink became the first and to this point only team from the 3-4 game to win the Brier since moving to the Page Play-off system in 1995.

Gushue has been knocking at the door for several years.  Last year his Newfoundland-Labrador squad ended the week in 1st place but failed to make the finals and had to participate in the inaugural Brier Bronze Medal game.  Perhaps he wants a rematch against Howard and a chance to replay the 7th end of the 2007 Brier finals, written up here:
Rob Fowler’s squad from Manitoba has big event experience as a team and some past Brier appearances as well with Stoughton (Rob) and Burtnyk (second Richard Daneault).  It appeared to be pre-destined that Mike McEwen would reach his first Brier and be a heavy favorite.  Fowler may be able to coast under the radar for a few days, but I expect them to challenge at the end of the week.

There are web sites out there that have Brier lines.  The site has some match-up bets for teams.  One I like is Koe to have a better record than Gushue as -140.  That’s roughly 58%.   Manitoba -130 against Northern Ontario is another attractive line.  Please wager responsibly and don’t consider any of my advice as an excuse to lose your hard earned money.

I will be “running back” to Saskatoon next Sunday to Thursday for the Brier.  Anyone who searches me out and buys me a beer will get more great gambling tips.


  1. Kevin,

    I don't see a ton of comments here so maybe this one will stand out, and you could provide a reply.

    I'm wondering if you know if there are game "charts" published for any major curling events. By "chart" I mean a shot by shot diagram of the in play area for each game. You have these diagrams within other posts, but judging by the graphics it looks like you made them yourself with the available curling software (a nice job). And I'm not too greedy - just an after picture, not a diagram of how the rocks moved with arrows.

    I know this was available during the Vancouver Olympics for each game, but they have disappeared as far as my searching can find.

    I have a very active interest in sabermetrics, with fangraphs being my favourite site.

    Curling stats are useful to the extent that you apply them in your writing, with the win percentage change after each end. However, I would like to study the rock by rock progressions to see if anything could be discerned from early, and middle, shot choices and the frequencies of end scores (-1, 0, 1, 2 etc.).

    I think it would also be a great way to look at an entire end at the same time for better analysis.

    It would great to hear either way if you know if these charts are published anywhere.


  2. I have written before that Curling organizations could easily use the existing sensor in the rocks (used for hog line violations), add a sensor into the ice at the pin (centre) and then have a computer program map every single shot. This would be useful for fans but extremely useful for analysis and analytics. But yes, I draw my own pretty pictures and have not seen any other tracking process like you mentioned. Perhaps for the next Olympics.