Saturday 23 February 2013

Swimming vs running: men vs women

In a older post, I took a ratio between men and women's world records in running for various distances. Top women run slower than top men, that much is clear, but taking this ratio I was looking for a pattern if we merge all the distances from 100 meters to 217 kilometres, obtained here, here, and here. Then I plotted the result as a log-graph. The x-axis isn't pretty, but I have an aversion to base-10 notation (I can't explain why).

Regardless of how I plotted the numbers, three orders of magnitude of data ought to reveal something. What you can see below is, with the single exception of the 100 km road race record is that as distances get longer the ratios increase. In other words women, compared to men, get slower for ultra-long races. Starting at the 60m dash Maurice Greene ran a 6.39 versus Irina Privalova's 6.92 for a ratio of 6.92/6.39 = 1.08, and ending with the 217 km Badwater Ultra that Valmir Nunes ran in 22h 51 min compared with Jamie Donaldson's 26:16 effort for a ratio of 1.15. Averaging the ratios for all the distances leads to an "expected" value of 1.12. To put that in perspective for middle distance people, we could say an 8:54-minute male 3000 meter runner is equivalent to a 10-minute female (which is why I'm annoyed at the women when 10 guys are running that fast and no girls, i.e. here). Consider that a male 4 minute miler is the same as a 4:30 female.

I imagine not many realize women are closer to men's performances for the shorter distances than for longer. Most semi-knowledgable runners will point out how close Paula Radcliffe's time is to top men in the marathon, but fail to realize the difference is even less in shorter distances (and the gap widens as you go farther). On explanation could be that short-distance women are more prone to drug enhancements.  Another is that men are more attracted to ultras leaving the women with a comparatively weaker field. Consider the field of the 2012 North Face Endurance Challenge 50 miler with 167 men vs 27 women. I find similar trends in other ultras. Very fast ultra women like Ann Trason don't change this general trend.

UPDATE: Curious, I decided to take male/female ratios of the CIS standards for indoor track and field. (If any non-Canadians are reading this, a university student running a "CIS standard time" at some point in the indoor season qualifies them automatically for the Canadian championships. Otherwise you must medal at the provincials). CIS standards are, sort of by definition, a time expected to be achievable by half a dozen or so Canadians. It is a good yard stick for the overall talent of each distance. Distances span 60m to 3000m (And I included the 4x800m relay to create a pseudo-3200m race). Here are the women/men ratios:


For all distances the ratios are well above the world-class averages. And it gets much worse with longer distances; the 3k ratio is 1.18. The average for all seven distances is 1.16. Keep in also mind that Canadian men are not themselves running incredibly fast times; the National men's 3000m can be won in the low 8 minute range, and the CIS record of 7:54 has stood for 30 years. This is not yet world competitive. The implications are clear: University Canadian women show a comparatively poor performance. I do not mean to pick on anyone, only point out an objective fact: Canadian university women are not running well. This is a fair statement as I'm comparing apples to apples. I also wrote about this earlier regarding the Canadian Olympic standards for men and women. These latter cutoffs hover around the 1.14 ratio mark, still lower than the CIS (i.e. so require the women to run faster).

The explanation for the slow women's CIS standards is not obvious. Female university enrolment is higher than the men's (most universities are at least 60% women), and women's track & field participation is at least as large as the men's (at McGill it was much larger). Also women's sports scholarships are not in short supply. Why are *specifically canadian* females falling short? At least several Canadian women should, for instance, be able to run 9:17 for the 3k or 4:16 for the 1500m (instead of 9:47 and 4:28). Megan Brown is the only Canadian who's close to these two marks.

Random shout-out to all: I would enjoy a coach's perspective on the issue.

(END OF UPDATE)

Now I wanted to compare this log-plot of running times with swimming. I've noticed a recent growth in longer distance swimming events growing in popularity. Consider for instance the 10km "marathon" swim introduced to the Olympics in 2008. Like running, swimming has distances ranging from 50 meters to many miles. I took a ratio of the men and women's records for pool races here (freestyle only). For longer distances I used records from the one and two crossing swims across the English Channel. The Channel is about 35 kilometres wide and a double crossing (there and back) took 16 hours 10 minutes for Philip Rush and Susie Maroney in 17:14 (both Australian). For a medium-long swim distance  (3.86 km), I chose the fastest exit swim times of the 2012 Hawaiian Ironman.


The opposite trend results in distance swim events as the running. In this case the fastest women are begin to match the best men. The ratios dip under 1.10 very quickly, i.e. all racing distance longer than 100m show less than a 10% difference between genders.

At more than 70 km/16 hours of swimming, however, the number of participants dwindle fast due to the dangers of drowning, so I didn't search for longer swims than this. That makes the women's times even more impressive, as there are so few competing. One hypothesis why women begin to catch men is that women have naturally high body fat percentages. But I wonder about this, as how difficult can it be for a man to gain body fat weight while still maintaining muscle. Not to mention men are taller swimmers which may add a possible advantage (certainly in a pool, anyway). The answer must be more complex.

Let me stop over-speculate. If anyone has a good explanation I'd love to hear about it.