Why Are Winter Sports Gender Segregated? Part 1: The Luge

I've written before about the absurdists claiming, without evidence, that men and women are perfectly identical on the athletic playing field — or would be if it wasn't for socially constructed limitations. This seems to come up every so often in areas where men and women play the same game, or at least similar games with the same name, with slightly watered-down rules for the women. Perhaps nowhere has this been more true than in tennis, as we saw this last summer when John McEnroe disrupted the zeitgeist by stating the obvious: a man at the peak of his game could whip a top woman without too much difficulty, claiming "if [Serena Williams] played the men’s circuit she’d be like 700 in the world"¹. This caused a good bit of predictable horrified tweeting from the chattering classes, but to her credit, Williams, who had embarrassed herself as a mouthy teen against #203-ranked Karsten Braasch, recognized the futility of a do-over:

“For me, men’s tennis and women’s tennis are completely, almost, two separate sports,” Williams said. “If I were to play Andy Murray, I would lose 6-0, 6-0 in five to six minutes, maybe 10 minutes. No, it’s true. It’s a completely different sport. The men are a lot faster and they serve harder, they hit harder, it’s just a different game. I love to play women’s tennis. I only want to play girls, because I don’t want to be embarrassed.”

But the case for sex segregation in the sports of the Winter Olympics would seem at least superficially different than tennis. Many of them are about going downhill very fast on waxed sticks or sleds, and as the acceleration due to gravity is 9.8 m/s² is true for everyone, it would appear those, at least, ought to negate most or even all of any sexual advantage. So, Maggie Mertens asks in Deadspin, why are winter sports sex-segregated? Why, for that matter, are there so many male sports for which there are no female analogues? Why, when men and women do participate in the same events, the course distances, and sometimes even rules are changed?

Luge

These questions need answering at some length. Mertens starts with the luge, which is maybe ideal for her argument. It's just a sled, so why is it that the men's Pyeongchang course (1,344.08m) is longer than the women's course (1,201.82m)?²

The sport in which you lay on a sled and hurtle yourself down an icy track the fastest wouldn’t immediately seem like it has any kind of bias favoring athletes of one gender or the other. But think again! The women competing in Pyeongchang will barrel down a track that’s 10.6 percent shorter than the men’s. That’s a difference of just 142m. And when it comes to doubles, women don’t have an event at all. Apparently only two dudes can lay on top of each other and fly down the ice on a sled.

I immediately grant the unfairness of having no women's double event, and possibly its absurdity, presuming there's no compelling reason to omit it. (There is a mixed doubles relay race.) I also concede the oddity of having a ladies start (reused for men's doubles). So why have sex segregation at all?

Looking at the final results page from the men's and women's luge events (both PDFs), doing the math and equalizing for course length, we find that gold medalist Natalie Geisenberger of Germany had an average speed of 93.4299 km/h, where her male counterpart, Austrian David Gleirscher, went a sizzling 101.492 km/h. That's 8.6% faster! Is the ice slicker for men?

One obvious answer might be that the initial push has a great deal to say about course times, and in fact if you compare the Gleirscher's fastest time (2.595s) with Geisenberger's (4.318s), you'll note that Geisenberger's is 66% longer than her male counterpart. Yet, at the end of the event, Geisenberger's average speed is 92% of Gleirscher's. Also, Gleirscher's averaged start time of 2.547s placed him in the bottom to middle of the pack of start times on each run; the consistently fastest luger at the start, Tucker West of the United States, who had two firsts, didn't even qualify for a fourth run. These in tandem strongly suggest the initial push isn't all that important, and/or we need to look more closely at why that difference might exist at all. So what else could account for it?

1. It appears that women have a longer distance to start their sleds, an artificial handicap. I base this on the fact that the men's doubles event (PDF), which begins at the women's start, has similar but shorter start times. (The shortest start time was recorded by the men's doubles gold medal winning team, Tobias Wendl of Germany, who clocked in at 4.174s, where the best time in the women's single, a tie between German Tatjana Huefner and Korean Aileen Kristina Frisch at 4.310s, is 96% of the men's speed. The average gold medal men's doubles course speed was 92.43 km/h, slightly slower than the women's single time.)
2. The track for the upper 142.86m of the men's course is steeper. I have no way of ascertaining this, and so ignore it.
3. Even if it is not steeper, the drop through the additional 142.86m provides enough boost to accelerate the men's sled to higher speeds. Using the average³ course 10% slope, we can work backwards to an estimated acceleration on an idealized course:
   
   
   A 10% slope is a rise:run of 1:10, or tan^-11/10 = 5.711°
   
   Acceleration due to gravity, adjusted for slope (ignoring⁴ friction due to ice), and assuming no other losses (as by non-ideal course traversal), we get
   
   a = 9.8 m/s² * sin(5.711°) = .9752 m/s²
   
   Distance is expressed by the equation
   
   x = 1/2at² + vt + x₀, where
   
   

   x = position in m
   a = acceleration in m/s²
   v = initial velocity in m/s (0 in this case)
   x₀ = initial position (also 0)

   This reduces to x = 1/2at², or transposing and substituting,
   
   t = sqrt(2*142.86 m/.9752 m/s²) = 17.12 s
   
   The final speed at the base for our idealized luger is thus
   
   v = .9752 m/s² * 17.12s * 3.6 km*s/m*hr = 60.09 km/h
   That's a pretty serious advantage! Of course, this is a back of the envelope guesstimate, without knowledge of the course details.
4. Wind speed on the course (if any). I also ignore this factor as evening out over time.
5. Strength and body control factor in steering and thus navigating the fastest path through the course.
6. Other, unknown factors.

The third element would appear to be, by far, the biggest factor. Looking through the results, you might notice a couple interesting things about the record keeping:

1. Average speed differs from finish line speed.
2. Fastest finish speed is not the same thing as a first place finish. German Andi Langenhan's second round finish speed of 130.5 km/h was a men's solo track best, but he only placed eighth with his time of 47.850.

 As this How Stuff Works article suggests, lugers have intense training regimens that would tend to confer advantages to men (as ever, emboldening is mine):

The start is the most important part of the race. It's the time when the slider is most in control, so his or her training can have the greatest affect on the outcome. Luge athletes build tremendous upper body strength for the start, when they'll propel themselves, their sled and any extra weights onto the course. Hand strength is also required for the start, when the slider paddles as quickly as possible for the first several feet of the course. Since a slider's body faces up to 5 Gs during a run, he must be in overall excellent physical and mental condition to manage the 50-second attack on his body and his focus.

In the summer months, luge athletes train hard to build upper body muscles through swimming, weight training and calisthenics. In the winter months, typical luge training includes practice runs every day. Sometimes, they'll practice only starts, developing strength, agility and technique.

So upper body strength — surprisingly — is the major focus of luge training, as is hand strength. And yet we see no compelling evidence that faster starts correlate to better finish times. As the luge is an event timed down to the third decimal (thanks to a .002 difference in the 1998 Nagano Olympics' womens' singles event, then within the margin of error of the timing system), superior male body control affecting steering could mean women might not end up on the podium very often. Yet at this point, I see no reason to rule it out entirely, either. While I see little reason to believe that male and female records will eventually match for sheer physical contests (in general, male and female records in summer Olympic games have reached a limit of about 90%), it is not obvious that this applies to the luge, at least.

---

¹ An interesting sidebar on Serena Williams' and John McEnroe's public throwdown comes in this Stats On The T post, which favorably compares Williams' serve speeds with those of top male competitors. But as the commenters following point out, service speed is only one part of the game, asserting (without specific data) Williams lacks the three-step sprint speeds of her male competitors. Given Williams' avowed refusal to play men, I take her as an authority.
² The governing body for luge events, the International Luge Federation, or FIL for its French acronym, Fédération de Internationale de Luge de Course, prescribes flexible course lengths in their posted rules (PDF); men's courses may be up to 1,350m, with a minimum of 1,000m for men's singles. All other courses must be no less than 800m, with no specified maximum; presumably the women's start is customarily set after the men's.
³ From FIL rules Supplement 1 §3.1.1, "The average gradient of a track from the men’s start to the low point should not exceed 10%." It could of course be lower, though in §3.1.1 of the general rules it says, "The start ramp should have a gradient of 20-25% and a length of min. 10m and max. 30 m." I don't know whether the start ramp is considered part of the overall time or not, but it seems reasonable that it would be. I have sent an inquiry to the US luge team's offices and await their response.
⁴ An earlier version of this post had an erroneous calculation for the acceleration due to gravity on the inclined plane that included friction, so I simplified it to ignore friction.