Running on Tilt
Data on NYC Marathon splits
When Pheidippides raced from Marathon to Athens 2,500 years ago, he ran without a D-Tag to record his split times. Given his death at the end of the run, one can assume he gave maximum effort, though without his splits it is impossible to judge whether he ran as quickly as he could have had he taken into account the wisdom of another ancient Greek, Aesop, who lived 2,600 years ago. What distance runner has not heard Aesop’s fable of the tortoise and the hare, and is not versed in its moral, that slow but steady wins the race?
Aesop’s lesson remains undigested by the masses. Despite marathon training involving countless hours of running, thousands of miles logged, feelings emoted through runner blogs and bottomless conversations with fellow runners, well over 9 in 10 New York City Marathon participants last fall ran more like the hare, slowing down over the course of the race, than the even-splitting tortoise.
If you take the top 100 finishers in each age group, 8.2% of women and a scant 3.4% of men ran negative splits. The percentages are likely lower for the entire field, as faster runners generally ran more evenly than slower ones.
The winners ran negative splits. Winner Gebre Gebremarian ran the first half in 1:05:20 and the second half in 1:02:54, for a negative split of 2:26. The first woman, Edna Kiplagat, ran a 3:16 negative split. Eight of the past twelve NYC marathon winners in both the men’s and women’s races have run negative splits. But these pros are the exceptions. The average breakdown for marathoners who ran sub 3 hours was about a 5-minute positive split, those who finished between 3 and 4 hours averaged about a 9-minute positive split, and those between 4 and 5 hours averaged a positive split of about 15 and a half minutes (these data stem from evenly distributed samples of runners within those time ranges).
Older runners were not wiser. Those most likely to post negative splits were the 25-29 year old runners for both men (9%) and women (20%). Of the top 100 men in the 50-54 year old age range, only one ran a negative split. There was a kink in the data. Runners who finished between 3:00 and 3:30 actually had a slightly higher positive split than those who finished between 3:30 and 4:00. This may be due to blow-ups among those trying to break three hours, and/or the relatively high number of women in the latter sample, as women ran more evenly than men.
The second half of the New York City marathon course has slightly more elevation gain than the first half, and the difference in terrain is estimated to cause a 1-2 minute positive split given evenly distributed effort. If one assumes that a runner would achieve his or her best time on any given day with an even effort, the data still suggest that runners ran inefficiently.
The 800-meter race is one in which world records have been set by positive splitters, but not the marathon. The three fastest marathon times ever posted resulted from negative splits. Of the world’s 28 fastest marathon times, 16 were run with a negative split.
The inability of the body to store enough glycogen for 26 miles of racing and the buildup of lactic acid are common explanations for why marathoners run positive splits. When I slow down in these races it is because my body is stiffening, not because I am gasping for breath. I posted an ugly positive split at the New York City marathon despite blogging about the very topic of pacing beforehand, and despite running with a track club that races marathons under the philosophy of running 10-seconds-slower per mile than the goal pace for the first 10 miles of the race. When I ran the race in 2006, I bonked to the tune of a 21 minute positive split. This time, I ran the first half in 1:28:43 and the second half in 1:39:29, for about an 11 minute positive split. My goal had been 2:55, and my secondary goal was 3:00. I finished in 3:08:12. In all other races from the half mile to the half marathon, I generally run even or negative splits.
If I had held back more at the start, would I have been able to break 2:55, or 3:00? I doubt it. Of the 938 runners who broke 3 hours for the marathon, including 95 who finished between 2:59:00 and 2:59:59, only five ran the first half of the race slower than 1:30. To run an evenly paced race, I probably would have had to take it out in 1:31 or 1:32.
Aesop’s simple fable is so hard to apply because the marathon is not so much a race as an anti-race. The way to run the fastest marathon is to run more slowly than you want to for a long time, and, of course, to be in great shape. After all the training, keeping that competitive urge in check requires bottomless patience. The silver lining to the massive failure to adhere to Aesop is that most of the 45,000+ finishers of the New York City Marathon could shave minutes off of their time by simply being more like the tortoise and less like the hare.