Thursday, 26 April 2012

AC women's Olympic cutoffs

It's Olympic season, so why not another Olympic-related post?

There has been some recent discussion whether Canada's two women Lanni Marchant and Krista Duchene (who ran 2:31:50 and 2:32:06 marathons at Rotterdam, respectively), should go to the Olympics. Why or why not? Their times were well under the 2:37 A standard set by the IOC, but slower than Canada's own standard of 2:29:55. Simple, yet complicated.

For a counter-example to these practices, consider admission into the California Institute of Technology, one of the more prestigious universities in the world. I might dare say it's harder to get in to Caltech than the Olympics given there are only 230 openings a year. How do they decide who gets in among the 5,200+ applications? From their webpage:
The decision-making process is much more of an art than a science here.  There are no formulas that assure success, and even for experienced application readers it is impossible to predict the outcome for a particular applicant without reading the entire application.
There are no automatic cutoffs for any criteria that we might consider, and every application submitted for admission to Caltech is read to get a good sense of who you are and whether you would be a good fit for the Caltech community.
And that's from a science-based institution. Conversely, too, there is no 'minimum' SAT score that grants automatic admission. Smart institutions know simple rules do not necessarily produce the results you want.

But here I am comparing apples and oranges. Smart students have nothing in common with dumb athletes who just have to run fast in a straight line, right? But if that's the case I suppose it's really easy to train fast athletes. There's nothing to it: set a high standard and away they go! Obviously not: it's the years of training and development that separate the excellent from the merely good. Hence why the Kenyan Olympic marathon squad was not handed to the three fastest runners, but instead chosen for a variety of reasons and by an experienced coach (Canova), not a spreadsheet. More recently Canada's Olympic squads have been chosen with the same care. Athletic development is not rocket science; it's much harder than that.

Choosing a cutoff

There's more to racing than just placing arbitrary hurdles in the way. Nevertheless, minimum standards are sometimes needed just to keep the competitors clustered. The question du jour is whether the cutoffs given by AC for women are in any way unfair. Below I've posted the international Olympic A and B standards.
Below is the world record list for both men and women for a variety of distances. (I have not included all of them in the next plot for clarity's sake). You'll see where I'm going with this in a minute.

Event Record (men) Record (women)
100 m  9.58 (+0.9 m/s) 10.49 (0.0 m/s)
200 m 19.19 (−0.3 m/s) 21.34 (+1.3 m/s)
400 m  43.18 47.60
800 m  1:41.0 1:53.3
1000 m 2:12.0 2:29.0
1500 m  3:26.0 3:50.5
Mile  3:43.1 4:12.6
2000 m 4:44.8 5:25.4
3000 m  7:20.7 8:06.1
3000 m SC 7:53.6 8:58.1
5000 m  12:37.4 14:11.2
10,000 m 26:17.5 29:31.8
10 km (road) 26:44 30:21
15 km (road) 41:13 46:28
20,000 m (track) 56:26.0 1:05:27
20 km (road) 55:21 1:02:36
Half marathon 58:23 1:05:50
25,000 m (track) 12:25.4 1:27:06
25 km (road) 1:11:50 1:19:53
30 km (road) 1:27:37 1:36:36
Marathon  2:03:38 2:15:25
100 km (road) 6:13:33 6:33:11
I'm going divide the record women's times by the men's and do the same for the Olympic A and Canadian A+ standards. This will create a set of comparative ratios for each distance:

As you can see there is considerable variation in both the two set standards and the WR times. Overall notice that (observing the blue bars) that women are about 10-15% slower than men in every event from the 100m dash to the marathon (the average for this collection is 12.2%: a ratio of 1.122). The ratio seems independent of distance run and includes a great deal of variation.

It is clear that the Olympic A ratio for both men and women's qualifying times is consistently higher than the WR ratio (listed for sports included in the Olympics). I calculate the women's A standard is on average 14.1% slower than the men's. The Canadian Olympic A+ standard is 13.1% slower, still more generous than the WR average. These set points are no fluke: The ratios imply both the IOC and Canada have specifically relaxed women's standards, quite deliberately, so that more women may participate in the games. Fewer women reach close to the same zenith of performance as men. These  cutoff marks avoid problems of unrealistic expectations or too few female competitors.

The cutoffs chosen by AC seem to be mostly fair, relatively speaking, except for the 5,000 m and the 100m, which by this graph are questionable. The marathon ratio for the A+ (14.1%) is less than the A level (16.3%) but not out of line with respect to other running events and certainly much slower than Paula Radcliffe's WR time. Clearly AC expects more from Canadian women than the Olympics does. Why, I'm not sure.

I dislike cutoffs as a replacement of personal judgment, yet I must admit women in Canada (and the rest of the world too) are not being treated unfairly compared with men's entrance standards. This data will probably not help Lanni or Krista in their bid to race in London, but I feel obliged not to fuzzy the big picture. As I pick my battles, this will not be one of them.

No comments:

Post a Comment