The increasingly parochial observations of a casual runner in his fifties. Was "serious" about "the sport" until personal and sociocultural inevitabilities prevailed.

Thursday, May 5, 2016

Not one for the ages: a strangely short-sighted NY Times effort

Gina Kolata has written a piece for the New York Times about a cold, recently unearthed fact: Runners slow down as they get older. This is arguably not her least praiseworthy effort for the Times just this spring alone, given what came one week later, but in any event, this bright journalist and passionate writer manages to again portray a science or health issue as something other than what it actually is.

In case you're feeling too lazy to follow a link, I will summarize the Times article in one sentence: A Yale economist and runner has produced a mathematical formula, translated into an online tool, that predicts exactly how much runners can be expected to slow with increasing age, and even more helpfully, how fast they would have run when they were younger if only they had tried.

I'm not quarreling so much with what Kolata writes in this piece as with the apparent knowledge gaps that allowed her to go forward with the story to begin with. She's been writing about endurance activities for a very long time, for a well-respected newspaper, and so I would expect her to be more thorough about doing her homework.


I see numerous problems with the article and its subject. One, such a tool already exists and has for years, in the form of the World Masters Athletics (WMA) age-grading calculator (that's one page of many like it, and possibly the most recent one). People take issue with some of the data that this calculator serves up, but when won't they? Runners are often the kinds of people who will scream holy hell when their Garmin tells them a certified 5K is really 3.14 miles while simultaneously claiming a time run on a course known to be short by a solid tenth of a mile as a personal best. For reasons I'll get to, though, this tool is undeniably better than the one the economist in the article, Ray Fair, came up with. And Kolata seems unaware of the whole WMA system.

Kolata also fails to note that Fair's model uses data only for men, She also, inanely if harmlessly enough, says that aging runners are forced to compromise, because if they run faster they can't do this for as long and if they go farther they have to slow down. Is there a difference? She also seems surprised at the idea that people's muscles start to fail with age even with heavy use, as if aging can simply be countered with a ample effort. The whole delivery has echoes of crankery, and is steeped in the apparent assumption that the writer is addressing thoroughgoing idiots.

Most of the problems I see, though, are with Fair's model itself. It assumes that runners as a rule are at their fastest at age 35 from the 800 meters to the marathon. Now, to be fair (you knew that was coming, right?), a model like this one requires some baseline assumptions, and in this case Fair is apparently assuming that they typical competitive runner, training at the same diligent if not obscenely dedicated level, is apt to peak in his or her mid-thirties and then start his or her inexorable slide. I would argue that most serious runners I know have at least an odds-on chance of running their best long-distance races, particularly the marathon, after age 35, because so many of them don't even start running seriously until they are in their thirties. This throws the entire curve of decline off. If you start setting personal records at age 40 thanks to a big uptick in training, and keep pressing hard, you're not going to find that age 45, you've undergone the predicted 10 years of steady unraveling.

As for running your best 800 meters or mile on the track at that age, that's a much iffier proposition. Recent history doesn't feature a lot of world or Olympic champions in track events, even up to the 10,000 meters, who are close to 35. I am probably being uncharitable in thinking mainly of unusually fast runners here, though,

Fair assumes a proportional slowdown across the spectrum of events. That is, if you lose five seconds off your 800-meter time between the ages of 35 and 40, you're expected to lose about 10 seconds a mile in the 5K, 10K, half-marathon, marathon and anything longer than that. I don't think this is most people's experience.

Finally, Fair muses that he model likely won't work as well for women because women tend to slow down with age more than men. This is almost assuredly wrong, as women as a rule can stay far closer to their best lifetime performances, and achieve a lifetime peak later, than men can.

If this were an article in a published periodical and not a blog post, I would have to back a lot of this up with hard data or at least find some qualified person to throw me some quotes to use, but it's not, so I don't. I will once again admit that Dr. Fair's model not applying well to elite runners and others close to the fast end of the performance curve has me more disdainful of the model than I otherwise might be, but even he says that he's disappointed because his own model doesn't apply to him. (His page with the running calculator has some other cool stuff on it, though.)

Anyway, a shorter version of this post and the Times article rolled into one: Yeah, you're gonna get slower, and you can live with it on its face or you can dig up calculators that soothe your ego using numerical handicapping schemes providing the right woulda-shoulda-couldas.




No comments:

Post a Comment