The Quadrennial Canada Olympic Trials Preview - Part 2

Some interesting analysis in this month's digital edition of The Curling News.  Terry Jones gave his odds for the event and, unless it's a typo it would appear Terry failed high school math.  He has Jacobs (9 to 5) as the higher favorite over Gushue (5 to 2).  These odds imply that Jacobs has a 35.71% chance of winning, while Gushue is 28.57%.  I think those Brad's should be switched.  Epping at 115-1 seems a tad high to me. Perhaps it is supposed to be 15-1.  If not, I'll gladly drop $5 on the Toronto squad with Mr. Jones, if he'swilling to accept my wager.

Looking at Terry's write-up on the womens, possibly another typo, he has Homan at 5-2 and Sweeting at 9-5.  9-2 would make more sense here (and with Jacobs).  Flaxey at 5-1 could be too generous, and the numbers below will explain why.

Reminder on the numbers.  Before each Olympic Trials I breakdown the results of each team against each other team.  I look at head to head wins and losses, scoring and then examine winning percentage for both historical (lifetime of the teams) and the last Olympic cycle (start of the 2014/15 Season).  I then look at a Combined result, which more heavily weighs the last cycle while still accounting for historical results.  Using Bill James log5 method and his Pythagorean expectation, I estimate the expected wins of each team.  Unless otherwise stated, I use the Combined results in the calculation for Expected Wins (EW). Keep in mind these sample sizes for womens is much smaller than mens, as these teams don't compete as often over the course of a season.  More variability in our prediction is the likely result.  

Women's Teams

In the mens bracket, 5 skips also appeared in the 2013 Trials in Winnipeg (Koe, Mcewen, Epping, Jacobs and Morris) and a 6th was nearly there (Brad Gushue lost the final Pre-Trials spot that season).  The womens draw this time around has only 4 skips from 2013 (Homan, Jones, Sweeting and Carey).  It will be interesting to see how these Trials rookies will handle the pressure and nerves of the event.  Some could falter, while others might embrace the moment, playing loose and making a deep run. 

The Favorites 

• Speaking of playing loose, how will local hero and reigning World Champion Rachel Homan handle this event?  All eyes (including those of Curling Canada) are expecting a victory, and sometimes expectations can be difficult to carry on your shoulders.  They handled the pressure well last year at the World Championships, I expect they can do the same next week.  It might surprise you to know this squad is only 65% against this field since the start of the 2014/15 season.  Englot (care of her Scotties performance) is 4-2 against Rachel, otherwise they hold a winning record against everyone.  Sweeting has 7 wins against them to 10 losses, and Scheidegger 4-5 may also be a tough out.  EW = 5.53.

• Like Brad Jacobs, Jennifer Jones is defending Canada Olympic Trials champs (does it feel like a defense when it was four years ago?).  Jones is 13-14 against Homan all-time but only 5-10 since the last Olympics season. They're 13-6 vs Sweeting during that time and have a surprising 57% winning percentage against this field.  That's a drop from their all-time winning percentage of 64%. EW = 5.32

The Contenders

• I'd like to think another team or two would make this grouping, but the numbers indicate otherwise.  Team Sweeting is 56% against this field including 7-0 against Englot.  In this last cycle, they have 23 losses from the two favorites, versus 13 wins.  Omit Homan and Jones from the data, and Sweeting is 75% against the rest of the field. EW = 4.7

The Challengers

• Teams Carey, McCarville and Englot all have a  44 to 45% winning percentage against this field the past 3.5 seasons.  Granted, Englot (15-19) and McCarville's (10-12) sample sizes are smaller, so it's unclear if these are a good indicator of likely results.  Other than Sweeting, Englot has shown she can take on the best teams (6-5 against Homan and Jones, 5-4 against Carey).  Chelsea Carey is only 3-10 against Homan and jumps to 50% against the field if we remove those games.  
• Scheidegger has only a 35% winning percentage against this field, but they've beaten Homan 4 times in 9 tries.  Their all-time percentage is 40% and recently have taken lumps from Carey and Sweeting (3-11 combined).  

• Expected Wins
• Carey = 3.67
• McCarville = 3.48
• Englot = 3.98
• Scheidegger = 3.29

I'd expect one and likely two of these teams to exceed these EWs and possibly land in the play-offs (or at least a tie-breaker).

The Underdogs

• Terry Jones had Team Flaxey at 5-1 odds.  With only 20 wins all-time against 40 lossess, and 13-28 since the last Olympics, the numbers would suggest 2.83 wins and no playoff appearance.  If you take out her 5-3 record against Tippin, Flaxey's winning percentage drops to 29%!  Two wins against Homan and Jones each means they can hang with anyone here, but their point differential vs this field of -1.77 per game is the worst of any team.  Homan, for example, is +1.43/game and Jones is +1.11, while all other teams are in the red.

• While we're talking point differential...in the mens (since I missed it in Part 1), Koe, McEwen, Jacobs and Carruthers are in the black, while Gushue is at -0.04 pts/game.   These are all-time numbers and I suspect a positive number for Gushue if we looked at the last 3+ years.  
• Womens games are actually a full 1.07 pts scored (total) more per game than the mens. Scheidegger at 10.65 pts/game is the lowest womens while Morris at 10.61 pts/game is the highest mens team.  I didn't have the time (or interest) to adjust for 8 vs 10 end games, so some skewing in the numbers (higher percentage of womens games are Scotties vs Brier in mens). 

• And finally....Team Tippin has very few games against this field.  They sit 9-16 all-time and 5-8 during this last Olympic cycle.  They have yet to face Sweeting and only a single loss against Scheidegger, but at least one win against every other team (including Homan and Jones) should give them some hope, right?