That Geoffrey Mutai won the Boston Marathon last year was no surprise. That he won in 2:03:02, however, is another matter. "Stunning" is a word often misused, but not in the case of the fastest marathon ever run. The fine points of why Mutai's time does not count as the world record have been oft repeated in the past year: Boston's course is point-to-point with a net elevation loss that exceeds the IAAF's limits, so is not considered "legal" for record purposes. Yet few in the sport consider Boston an easy course, with its early and sharp downhills known for coming back to haunt, if not cripple, incautious runners later in the race. Isn't there a loophole somewhere under which an otherwise challenging course such as Boston's or New York's could—and should— be made record-legal? Perhaps the wind should be treated like a tailwind in the 100-meter dash: If it's below a certain limit, the finishing times count for records. Or a marathon could get extra credit if it doesn't use pacers—who, after all, are there to lead the winner to a faster time than he could run on his own. Amby Burfoot, the former editor-in-chief of Runner's World Magazine, and David Powell, a former Olympic writer for the London Times, have their say.
Amby Burfoot: In April 1968, I won the Boston Marathon in 2:22:17. Eight months later, I ran 2:14:29 in the Fukuoka Marathon in Japan. Is Boston an aided course? You certainly wouldn't think so from my results in 1968. I ran nearly eight minutes faster in Japan, a huge difference. And yet I have never had any doubt that Boston is, under perfect conditions, an aided course—and therefore one that can't be eligible for record-setting performances.
Last April we saw why. Under near-perfect conditions—cool, dry weather, and a consistent tailwind—Geoffrey Mutai and Moses Mosop ran almost a minute faster than Haile Gebrselassie's [then-]marathon world record. Gebrselassie has been universally lauded as one of the greatest distance runners of all time, from the mile all the way up to his marathon record of 2:03:59. At Boston, Mutai and Mosop ran 2:03:02 and 2:03:06. How could they be that much faster than Gebrselassie?
Answer: They couldn't be, and they aren't. They just happened to benefit from racing on a course that perhaps one year out of 20 is too fast to be considered record-eligible. Heartbreak Hill? Yes, it's there every year. But so is Boston's net elevation drop of 477 feet.
David Powell: Wow, I tip my hat. I certainly cannot compete with Amby's performance credentials. However, bearing in mind his comments, I do find myself wondering whether I could have spent the last 27 years claiming a PR of 2:25, maybe 2:24, had I enjoyed the advantages that Boston offered last year rather than those dealt me by London in 1985 when I clocked 2:26:48.
But enough of us. I fail to see why Boston would be too fast to be considered record-eligible only one year out of 20 and not every year, when the elevation drop is more than three times that permitted by the IAAF. I hear the argument about Boston's sharp downhills but the overall net-drop rule is the only workable one and Boston, I'm afraid, must live with that.
Which brings me to the center of this debate, the other reason why Boston does not meet IAAF records criteria: It falls outside the rule requiring the separation between start and finish to be no more than 50 percent of the race distance. Runners can benefit unreasonably from tailwinds on point-to-point courses, as Mutai and Mosop did, and Amby's comparison of these two with Gebrselassie suggests an assistance of at least one minute, with the wind gain greater than the downhill advantage.
AB: This is probably a good time for me to backtrack just a bit. I don't want anyone to get the impression that I think Boston is an easy course. I don't. Most years it will produce performances that are slower than those on the other big (and flat, and record-eligible) marathon courses like London, Berlin, and Chicago.
Even with the net elevation drop, Boston's uphills (and steep downhills) are more than enough to do serious damage to the marathoner's leg muscles. After 20-plus miles, that damage changes your stride mechanics. Your body screams for relief, all the way up through your hips and your back, and your legs begin to shuffle like those of an 80-year-old. You hurt all over. You slow down.
Yes, this happens any time you run 26.2 miles. But few other major marathon courses punish you as much as Boston.
Nonetheless, every so often the weather gods line up perfectly with the course profile to produce a special situation where super-fast times are possible. It doesn't make any sense to accept these times as world-record performances. It would be like accepting 100-meter times with the currently allowable 2.0 meters/second tailwind and ALSO a 2 percent downhill track.
DP: You mention, Amby, the 2.0m/s tailwind allowed for 100-meter times. The question has been asked whether the same allowance might be applied to point-to-point marathons. The answer is surely "no." It would be impractical and complicate the sport, but I'll come to that later.
The average tailwind rule is, in itself, arbitrary, dating back to an IAAF Congress decision in 1936 based on studies from hand-timed races. In 1977 the IAAF required performances to be measured to 0.01 seconds using automatic timing but the 2.0 m/s limit has been retained for continuity of statistics.
How much faster would 2.0 m/s of following wind make over a marathon compared with, say, still conditions? Studies into the differences over 100 meters have shown that the advantage is 0.10 seconds for world-class men and 0.12 seconds for world-class women. By a process of simple mathematics that would mean, over the marathon, 42 seconds for a man and 50 for a woman.
It is not that straightforward, of course, as the slower the running, the greater the time differential, so the gain for an elite marathoner would be minutes. Here's one for you, Amby, a bit of fun. Take an instinctive guess, don't work it out. If a world-class male 100-meter runner continued at 9.90 seconds for the marathon distance, what would be his time? And a world-class woman at 10.90 seconds?
AB: David, I think your "simple mathematics" exercise yielded reasonable answers with regard to the potential Boston Marathon advantage—let's say 40 to 60 seconds. But I don't understand the sprinter-to-marathoner challenge. We all know that sprinters suck at the marathon, and marathoners suck at the 100-meters (or otherwise we would all surely pick the shorter distance as our primary event.)
Here's what's important: The Boston Marathon (and quite a few other marathons that don't happen to attract world-class fields) has a point-to-point, downhill course that under certain conditions make the Boston course faster than a flat, loop course. For that reason, Boston performances can't and shouldn't be considered for world-record purposes.
This does not make Boston a lesser marathon. Indeed, no other marathon comes close to matching Boston's glorious history. It doesn't mean that Boston's organizers should change the traditional course to a downtown loop that would make it record-eligible. It doesn't mean that runners should skip Boston every April and focus their attention on the Flatland Marathon.
Instead, we should treasure the Boston Marathon for all that it is, has been, and hopefully will remain. Lastly, we should acknowledge that road records can never meet the same strict standards as track records. Most tracks are largely similar, with only small distances. Road courses are distinctly and uniquely different, unified only by distance. I say: Vive la difference!
DP: Hear, hear, and let me add how impractical it would be to attempt taking wind readings along marathons courses, not to mention costly and labor intensive. How many wind gauges would be needed? Where would we place them? How misleading might it be to work out an average over two to two-and-half hours?
It is not as straightforward as track regulations, which require gauges to be positioned 1.22 meters above the ground and not more than 2 meters from the track. In 100-meter races, for example, wind velocity is averaged over 10 seconds from the start. Complicating the matter further, it has long been recognized that the presence of grandstands, or other buildings, may cause erratic wind patterns.
The marathon race is complicated enough without messing with it further. As well as those issues touched on already—downhill and point-to-point courses—we have a plethora of others, such as the use of pacemakers and mixed or women-only races counting for records.
Incidentally, I was not intending to suggest that a 100-meter runner could maintain speed for a marathon. That would be absurd. It was a hypothetical question for curiosity's sake. Mildly interesting, I thought, to know that the speed at which a world-class 9.90 seconds 100 meter male travels converts to just inside the hour for a marathon. For a woman running 10.90 seconds, it's around 1:16.
AB: Our editor asks whether a phalanx of pacemakers on a flat course like Berlin is the equivalent of a tailwind in Boston. Many top-level marathons (but not Boston or New York) attempt to produce world-record performances by utilizing pacers to assist the expected winners. I believe this is modestly helpful. But it's completely acceptable per IAAF regulations in track and road races, and I don't have any problems with it.
In addition, pacers aren't nearly as important in running as they are in cycling. Tailwinds and downhill courses have a bigger impact.
DP: Our editor further asks why it is legal to have pacemakers to help runners go faster but not legal to have a tailwind. The answer is one of equal opportunity. Pacemakers can be used by any race; the fact that some do not is their choice. Tailwinds benefit only those races with point-to-point courses. Loop courses do not have the option to use a tailwind.