Thursday, November 26, 2009

Credit Where Credit is Due: American and Japanese Men Aged 18-22 pt. I

by Brett Larner

special thanks to Ken Young at ARRS, the IAAF and All-Time Athletics for their database assistance in preparing this article

Last month I published a comparison of the results from the American NCAA Pre-Nationals XC Meet and the Hakone Ekiden Qualifier Road Race which showed that more Japanese university runners were running as fast or faster for a hilly 20 km on the roads than American university runners were running for a hilly 8 km XC. I received a fair amount of response to this comparison, much of it negative and much of it from American university runners, in the comment section, on message boards such as letsrun.com, and in my email inbox. One such letsrun poster asked what was apparently supposed to be a rhetorical question:

So how many of these "Rising Sons" have run sub-13:30 at age 18?

Yeah.
The poster was referring to Americans German Fernandez and Chris Derrick, both of whom achieved this impressive feat in late spring this year, clocking times of 13:25.46 and 13:29.98 to become the first American 18 year-olds to break this barrier. At the time I didn't know the answer, but it struck me as a good question. I recently had the time to look into it and found, unsurprisingly, perhaps, that the answer is 'zero.' Japan's best current track runner, 22 year-old Yuki Sato (Team Nissin Shokuhin) came the closest, clocking 13:31.72 at age 18. Very close, but no cigar. Were Sato an American this time would put him at all-time #3 well behind Fernandez, just behind Derrick, and far ahead of star Galen Rupp's 13:37.91 mark, but the letsrun poster is correct, the U.S. has two men with such times while Japan doesn't.

Thinking some more about this, I wondered how the numbers would look if you expanded the question out from its original one-dimensional form to include data for 19 year olds, 20 year olds and other comparable ages and distances. I compared runners from the U.S.A. and Japan at five ages from 18-22 and looked at results for 5000 m, 10000 m, half-marathon and marathon, even though the latter is rarely raced at such ages in either country. Setting the letsrun poster's arbitrary criterion of sub-13:30 as a benchmark, I used the popular McMillan calculator to find comparable marks for these longer distances. The McMillan calculator gives times of sub-28:03 for 10000 m, sub-1:02:24 for half-marathon, and sub-2:11:36 for the marathon as equivalent to a sub-13:30 5000 m. The results are quite interesting and seem to show clear trends in where the emphasis in coaching, racing, ability or some combination of all three lie.

The table below shows the complete all-time listings by age and distance for runners from the States and Japan who have achieved these times. With the exception of the marathon, in cases where no runners of a given age have achieved the mark, such as Sato at age 18 in the 5000 m, the fastest mark is listed. Click image for a full-sized version. You might need to enlarge it a bit.

Among 18 year-olds, German and Derrick have set new standards in the 5000 m and will hopefully continue to do so as they grow older. Looking at the the 10000 m and half-marathon, only Akinobu Murasawa's winning time at the 2009 Hakone Ekiden Qualifier is of comparable quality, but Japanese athletes have overall performed at a higher level with their advantage increasing with distance.

At age 19 Japanese runners surpass their American counterparts at every distance. Each country has produced two sub-13:30 athletes, but Japan's Kensuke Takezawa and Yuki Sato each recorded times around 5 seconds better than America's Dathan Ritzenhein and Hassan Mead. In the 10000 m two Japanese runners are under the 28:03 mark, with Ryuji Ono's 27:59.32 nearly 25 seconds ahead of top American Galen Rupp. The half-marathon is similar, two Japanese 19 year-olds coming close to breaking 1:02 while the best U.S. runner clocked 1:04:32.

Age 20 is where things start to diverge and thereby become more interesting. American runners show notable development in the 5000 m, with five men going sub-13:30. Only one Japanese 20 year-old, Kensuke Takezawa, has achieved the same mark. His time of 13:19.00 is considerably better than top American Bob Kennedy's 13:22.17 but there is no question that the U.S. excels in depth at this age. Looking at the longer distances the States begins to show development in the 10000 m, with two runners under 28:03 led by Galen Rupp's sensational 27:33.48. Japan likewise shows progress in the 10000 m, five men going under 28:03 topped by Yu Mitsuya's 27:41.10. Japan's progress in the half-marathon is identical, again five men going under 1:02:24 with top 20 year-old Masato Kihara's 1:01:50 over two minutes ahead of top American Herb Wills' 1:03:58.

The trends of rapidly-growing depth in the 5000 m and gradual progress in distance for Americans and of a shift in emphasis to the half-marathon and beyond for Japanese runners continue unabated at age 21. While there is roughly equal quantity and, with the exception of Dathan Ritzenhein's 27:38.50, quality in the 10000 m, the U.S. dominates Japan in the 5000 m, ten men to three and Bill McChesney's 13:18.6 leading Toshinari Takaoka's 13:20.43. The converse is true in the half-marathon, ten Japanese men led by Hidemori Noguchi's 1:01:55 against lone American Scott Bauhs' 1:03:04. Japanese 21 year-olds even take the occasional swing at the marathon. Masakazu Fujiwara's 2:08:12 list-topper would be a noteworthy result at any age, but two more Japanese men have broken 2:10 at age 21, well below the 2:11:36 cutoff.

The trends again continue at age 22. Seventeen American men have broken 13:30 at this age including Galen Rupp's 13:18.12 indoor mark, and Alan Webb's 13:10.86 is the fastest on the chart. While the U.S. numbers in the 10000 m do not improve in quality, the depth shows progression with six athletes cracking 28:03. The gradual move toward longer distances continues as one American 22 year-old, Todd Williams, has broken 1:02:24 in the half-marathon. Japan's numbers in the 5000 m and 10000 m remain relatively constant with the exception of Yuki Sato's 27:38.25 just over a second shy of top American Rupp. The Japanese depth in the half-marathon likewise remains constant at ten runners under 1:02:24, but there is a noticeable improvement in quality as four 22 year-olds have broken 1:02. The reverse is true in the marathon where the quality tops out with another 2:08, this one from Muneyuki Ojima, but the depth has improved to eight men under 2:11:36.

So, in summary, prior to Fernandez and Derrick Japanese teenagers consistently performed at a higher level than their American counterparts from 5000 m up, shifting in focus to longer distances as they entered their 20's while Americans continued to improve their speed at shorter distances and exceeded the achievements of Japanese men at such distances. Not exactly a secret, but those are the numbers. Fernandez and Derrick are clearly talented athletes who may well rewrite the lists. In a follow-up article later this weekend I will look at some of the reasons for these trends and what they might mean for both countries' young runners.

(c) 2009 Brett Larner
all rights reserved

13 comments:

Anonymous said...

Quite an interesting read. Good job for ploughing through the data.

As a side note, i've been following the debate the entire time (on letsrun and here). Needless to say, the response was predictable and disappointing.

I do think it would be extremely narrow-minded for people not to respect each country's respective strengths. Clearly as you have pointed out, the americans are much deeper over shorter distances but the japanese are far better at the longer distances.

Personally im just glad that both countries are working hard to challenge the africans. It's much more refreshing to see different faces.

Oh, and i'm not from either country, so no accusation of biasness here :) looking forward to your follow-up!

Anonymous said...

Thanks for that, very interesting.

I also have no bias as I am British, but in my opinion the American approach is favourable. I feel you need to run fast at 5000m and 10000m first, so that you can push up your potential in the marathon. I think it is unlikely that you will be able to run in the 2.05 region without being able to run very close to 13 minutes for 5000m, and once you have started marathon training going back to shorter distances is always going to be harder, so why not fulfil your potential on the track first?

I also think that your stats highlight the differences in physiology and running technique between American and Japanese runners. The americans tend to be taller with a longer stride and greater speed, whilst the Japanese are generally shorter, with a running gait better suited to the marathon.

Thanks again.

Roberto said...

Excellent research.

Anonymous 2 is obviously correct in that if you want to get to the top rung of the ladder, you need the basic speed to 1) be able to run the pace comfortably, and 2) be able to find something in the final 400. And that's where the Japanese model falls down, with its singleminded focus on mileage (a strategy that indisputably works, to a point).

That, of course (and as I'm sure you'll address in your follow-up), is the reason for the divergence in success at longer distances between young Japanese and American runners. I'd be surprised if on average Japanese were not running at least 50 percent more miles per week. [The equation changes for women at the marathon distance, with less competition and lighter body weight factors in the ability of Japanese (and the few Chinese whom we don't suspect of blatant doping or hermaphroditism) women to run the enormous number of training miles that I believe has enabled them to perform disproportionately well on the world stage.]

Last point: if you go down to the 1500 and 800, and include senior runners of all ages, you see how the Japanese emphasis on overdistance shapes its athletic performance. Japan's national 800 and 1500 records are laughable in comparison with its half marathon (long ekiden distance) and marathon marks. Because few people are training seriously (and single-mindedly) for those short track events. And because the post-collegiate (i.e. corporate) support infrastructure is uninterested in short middle distance runners.

Simon said...

Thanks for the interesting read Brett and for doing some thorough research.

A small point but it's perhaps not too surprising that you received a fair amount of negative feedback when your Letsrun post was provocatively entitled "NCAA is nice and all..." Given the board's audience, this was something of a red rag to a bull! Nevertheless, I watched the Hakone qualifier and the university runners were damn impressive and deserved much credit.

@Anonymous 2 - Not sure I agree entirely with the assumption that runners with ambitions to compete in the marathon should all work their way up through the shorter distances on the track. The natural progression doesn't always work like that and the Japanese have been more competitive at major championships on the road than their American counterparts. Check out the 5000m times of the 22 y/o Americans on the table and then think about their relative success, or lack thereof, at the marathon distance. To use a recent comparison, Atsushi Sato finished 44 secs ahead of Dathan Ritzenhein in London earlier this year. On the other hand, it was sad to see Shimizu get passed by Martinez and two Portugese at the WC this summer with less than 400m to go and have absolutely nothing to offer in terms of a finished kick. There is clearly need for the Japanese to work on pace-changing and their finishing kick.

Anonymous said...

Thanks for doing all that research! Great read (and interesting comments, too).

I guess it`s too obvious for anybody to mention, but it matters when comparing two countries: the US population is about three times of that of Japan. On the other hand running is by far more popular in Japan than in the US I suppose.
So to me it`s interesting that all factors combined both countries are about equally strong overall providing for an interesting comparison.

Joe

Nate Jenkins said...

It seems as though the high level of performances by the Japanese at a younger age could be tied to there higher mileage high school systems. In the USA most of our HS programs are lost somewhere in 1950's training theory with 1 to 2 races and 2 to 3 interval sessions each week with very low mileage, 30 to 50 miles per week. I for one believe if we had a higher mileage more sensible base training focused jr. programs that we would be much much faster across the board in the jr. ranks. I also think the Japaneses extreme focus on the marathon and on very high mileage and very long tempo really hurts there progression on the track as adults. That said as a 14:04 5k/2:14 marathon guy I really can't cast too many stones at them for struggling in this area.
I'm very interested to see your next post on the issue and how much you feel training plays a part.
nate

zatopek said...

I think you also have to take into account how often the different races are run. In the US, half-marathon and marathon races are rarely run by that age group so I don't know if it is a fair comparison.

Anonymous said...

This is a perfect example of the American system being focued on maximizing aerobic capacity (VO2 Max) at a young age while the Japanese system focuses more on maximumizing aerobic efficiency (mileage, LT, AT).

Thanks for bringing this to light.

- MPR

Nobby said...

Very interesting stuff; thanks, Brett. If you want to go even deeper, any info on current runner? I mean, in the case of, say, Atsushi Sato. He's still current (still competing) so his record at 20 or 21 would be fine but when it comes to likes of Seiichiro Sasaki, he was competing in the 1960s. For US 10000m for 22 years old, except for Rupp, they are all long-retired. Should we count those guys? What do you think?

Adeel said...

Great post. This blog is a fantastic read on its own because the Japanese running scene is interesting as it is. Still, to create this sort of discussion with people like Nobby and Nate makes it even better.

Anonymous2 makes a good point about the age of marathon specialization, which is a very interesting point of comparison between the two countries (zatopek alludes to this as well).

I presume that the Japanese approach to the marathon is to specialize around 22, given the explosion in sub-2:11s at that age despite the fact that there's no shortage of 62-minute half marathon runners. I also presume that this comes after a few years spent honing speed at shorter distances.

Early specialization in the marathon is one way of explaining the recent surge in fast marathons by East Africans. Many of them are very young, chief among them Sammy Wanjiru, who debuted at 21.

This year alone, James Kwambai ran 2:04 at 26, Tsegay Kebede ran 2:05 at 22, Vincent Kipruto ran 2:05 at 21, beating out 20-year-old Bazu Worku who ran 2:06, Patrick Makau debuted with a 2:06 at 24, Gilbert Kirwa ran a 2:06 at 24, Yemane Tsegay ran 2:06 (4th at WCh) at 24, and Robert Kiprono Cheruiyot ran 2:06 at 21.

Compare that with, say, ten years ago, when only a handful of runners under 25 ran the marathon. I think the Japanese path of early specialization in the marathon is part of their success. Their inability to handle changes in pace is maybe a matter more of training than not having run enough track races early on.

jobs said...

Nice analysis. I enjoyed reading it.

About the ages, how did you handle boundary cases? E.g., if runner A was 18 yrs 364 days old and runner B was 19 yrs 1 day old, would they still be in different groupings? (I realize that this is a lot of work.)

As a follow-up, maybe you could do the US vs. Japan analysis for the female runners?

Bruce said...

Brett - I go along with your observations. What you have seen for collegiate ages proceeds from what happened in high school. In the 1500m to 5000m range, both genders, there is an astounding reversal of USA teen dominance in sprints and field events. You can see this in a PDF I put together comparing the 2 nations` high school athletes, based on published lists of top marks in 2008. I recognize limitations of my survey, since USA statistics are more subject to incompleteness, and a track 5000m is rare for high school students in the USA.

http://harrier.kantoxc.com/Reports/USA-Japan_HS_Elite.pdf

What I want to know is why and how.

And we really need some research on whether these young athletes go on to enjoy their sport beyond high school. I am troubled when I encounter Japanese college students in my classes who no longer run despite posting what I considered phenomenal times in high school, times that would put them on the top 5 of most USA college cross country teams.

Anonymous said...

The only thing I can think of to back up the Americans arguement is that most collegiate athletes do not run road races such as the half and full marathon while in college. Otherwise I love this article!!