Training Log Archive: barbIn the 7 days ending Apr 8, 2013:
| ||||||||||||||||||||||||||||||||||||||||
| « | » |
| » now | ||||||||||
|
| ||||||||||||
| Tu | We | Th | Fr | Sa | Su | Mo | |||||||
Sunday Apr 7, 2013 #
Note
I see that Valerie has already updated the junior-rankings.org site for us, through yesterday, which is awesome.
In looking to see whether some more F20 wins might push Isabel above 75 so she wouldn't have to petition (they don't), I investigated further into what was setting the overall level of the green rankings, using the April 6 version. The relative rankings (between races, and between racers) are determined by normalization based on overlapping people attending the different races. The absolute level of the rankings, however, is determined by the average of the overall rankings of the top three people across all the races. For green, those 3 people are Joe Brautigam, Will Hubsch and Andras Revesz. Interestingly, Joe Brautigam alone, at a score of 110.57, is responsible for an upward shift of over 6 points. (If he is taken out of the level-setting calculation, everyone's score goes up by 6.22 points.) Without Joe, Isabel would be above 75 and not need to petition. (And yes, I know that removing him would change all the *relative* numbers a little, but the system is designed to be robust in that way, so it would not be a lot. It is the level-setting of 100 for the top 3 that makes the big absolute difference in the numbers.)
My point is just that the absolute value of the rankings is not a reliable or steady measure of ability, and is highly dependent on who is at the top. And therefore, an arbitrary fixed numerical threshold for limiting who goes to JWOC (it has been 75 for the past few years) is a questionable approach for OUSA.
Interestingly, Joe's score of 110.57 includes a score for one race of 123.36 (NAOC Middle), a race in which the average of the top 3 runners (Nils Gumaelius, Joe and Martin Sacher) was 120.78.
Even in smaller ways, the girls are penalized. At the NAOC sprints, for example, green was split into Green X and Green Y because of the number of participants. A pace of 7:44 got you a race score of 80 in the men's Green X. But the same pace got you a race score of 76.13 in the women's Green Y. Presumably the Green X and Green Y courses were fairly comparable in terms of difficulty at the same event, so this points to the fact that the assumptions of adequate consistency and sufficient numbers behind the rankings are not being met.
I do think it is important to aspire to international standards. At our meeting the other day, Erin said that men's winning times at JWOC on the long are 4 min/km for 18 km, and we are not even close. But I'm glad that we're moving toward inclusiveness for girls going to JWOC, because I think closing off the pipeline at that stage hurts us in recruiting and motivating girls at all levels.
In the table below, "I Place" means Isabel's JWOC USA (F20) placing in that race. (Anytime it's a "2", Meg was racing...) Average score of the top three finishers (often men in M50) in that race is in the next column, and then Isabel's score for that race. Then Isabel's score relative to the mean of the top three finishers. Then the score Izzy would have had without Joe.
Another very notable thing about this table is the huge jump (10-15 points) between before and after JWOC (July, 2012). Going to JWOC, training with the team and Erin for a week, and then a week of competition -- this made a huge difference in Isabel's orienteering abilities. We should be providing that opportunity to as many girls as we can.
Note: I haven't taken the time to totally double check all the numbers.
In looking to see whether some more F20 wins might push Isabel above 75 so she wouldn't have to petition (they don't), I investigated further into what was setting the overall level of the green rankings, using the April 6 version. The relative rankings (between races, and between racers) are determined by normalization based on overlapping people attending the different races. The absolute level of the rankings, however, is determined by the average of the overall rankings of the top three people across all the races. For green, those 3 people are Joe Brautigam, Will Hubsch and Andras Revesz. Interestingly, Joe Brautigam alone, at a score of 110.57, is responsible for an upward shift of over 6 points. (If he is taken out of the level-setting calculation, everyone's score goes up by 6.22 points.) Without Joe, Isabel would be above 75 and not need to petition. (And yes, I know that removing him would change all the *relative* numbers a little, but the system is designed to be robust in that way, so it would not be a lot. It is the level-setting of 100 for the top 3 that makes the big absolute difference in the numbers.)
My point is just that the absolute value of the rankings is not a reliable or steady measure of ability, and is highly dependent on who is at the top. And therefore, an arbitrary fixed numerical threshold for limiting who goes to JWOC (it has been 75 for the past few years) is a questionable approach for OUSA.
Interestingly, Joe's score of 110.57 includes a score for one race of 123.36 (NAOC Middle), a race in which the average of the top 3 runners (Nils Gumaelius, Joe and Martin Sacher) was 120.78.
Even in smaller ways, the girls are penalized. At the NAOC sprints, for example, green was split into Green X and Green Y because of the number of participants. A pace of 7:44 got you a race score of 80 in the men's Green X. But the same pace got you a race score of 76.13 in the women's Green Y. Presumably the Green X and Green Y courses were fairly comparable in terms of difficulty at the same event, so this points to the fact that the assumptions of adequate consistency and sufficient numbers behind the rankings are not being met.
I do think it is important to aspire to international standards. At our meeting the other day, Erin said that men's winning times at JWOC on the long are 4 min/km for 18 km, and we are not even close. But I'm glad that we're moving toward inclusiveness for girls going to JWOC, because I think closing off the pipeline at that stage hurts us in recruiting and motivating girls at all levels.
In the table below, "I Place" means Isabel's JWOC USA (F20) placing in that race. (Anytime it's a "2", Meg was racing...) Average score of the top three finishers (often men in M50) in that race is in the next column, and then Isabel's score for that race. Then Isabel's score relative to the mean of the top three finishers. Then the score Izzy would have had without Joe.
Another very notable thing about this table is the huge jump (10-15 points) between before and after JWOC (July, 2012). Going to JWOC, training with the team and Erin for a week, and then a week of competition -- this made a huge difference in Isabel's orienteering abilities. We should be providing that opportunity to as many girls as we can.
| Event | I Place | Top Three | Isabel | Isabel relative | Isabel sans Joe |
|---|---|---|---|---|---|
| Flying Pig 4/6/13 | 1 | 80.91 | 67.85 | 83.86 | 74.15 |
| Flying Pig 4/5/13 | 1 | 70.88 | 67.76 | 95.59 | 74.06 |
| ISIC Champs 3/10/13 | 1 | 71.75 | 73.56 | 102.52 | 79.86 |
| ISIC Champs 3/9/13 | 1 | 63.78 | 71.97 | 112.85 | 78.27 |
| GNC 1/20/13 | 1 | 75.67 | 63.55 | 83.99 | 69.85 |
| GNC 1/19/13 | 1 | 81.54 | 68.97 | 84.58 | 75.27 |
| UltraLong Champs 12/2/12 | 1 | 85.75 | 66.78 | 77.87 | 73.08 |
| NAOC Long 10/20/12 | 2 | 84.67 | 69.35 | 81.89 | 75.65 |
| NAOC Middle | 2 | 81.54 | 68.97 | 84.58 | 75.27 |
| Western States 9/2/12 | 1 | 86.22 | 65.13 | 75.54 | 71.43 |
| Western States 9/1/12 | 1 | 85.13 | 64.3 | 75.53 | 70.6 |
| Western Mass 5/26/12 | 2 | 74.79 | 51.4 | 68.73 | 57.7 |
| West Point 5/6/12 | 1 | 77.81 | 52.78 | 67.83 | 59.08 |
| West Point 5/5/12 | 2 | 75.52 | 54.16 | 71.72 | 60.46 |
| West Point 5/5/12 | 2 | 70.52 | 63.03 | 89.38 | 69.33 |
Note: I haven't taken the time to totally double check all the numbers.
C • Otoh 42
Orienteering 1:00:00 [1]
"Hung" controls for the Prospect Hill meet.
Tuesday Apr 2, 2013 #
Orienteering race 2:23:00 [3]