How to poorly evaluate the effect of superspikes in one easy tweet
Pretending that shoe technology has altogether invalidated times on the track and roads is the lazy way out of trying to discern its impact
A week ago, Sally Bergesen, the CEO of Oiselle, expressed an idea about “superspikes” that contained one error of fact and failed to account for at least one critical variable, yet was received enthusiastically by various current and former pro runners. As such, the idea was perfect for its environment.
The March 9 Deseret News article Bergesen links to notes the explosive increase between the 2021 and 2022 NCAA indoor track seasons in the number of athletes running under given time barriers, such as 13:50 for the men’s 5,000 meters. Two and a half weeks before Doug Robinson filed that story, I wrote a post predicting that it would soon appear, noting the rise over the past decade in the number of collegiate runners breaking selected time barriers each season—for example, 13:50 for the men’s indoor 5,000 meters.
Robinson allows that some of this increase is likely owed to NCAA athletes being granted an extra year of eligibility owing to COVID scuttling their collegiate progress. He also attempts to reconcile the Dragonfly ZoomX shoes becoming officially available in late 2020—he dates their arrival at “a few years” ago—with the 2021-to-2022 bonanza by observing that superspike availability among NCAA athletes is only now approaching unity, an explanation I accept.
It’s unclear whether Bergesen has accounted for these same confounding variables. But neither she nor Robinson seems aware that the 2021 NCAA indoor season was an anomaly for the glaring reason of the “2020” NCAA Cross-Country Championships being held in Oklahoma in March of 2021, two days after the 2021 NCAA Indoor Track and Field Championships were wrapped up in nearby Arkansas. (At least athletes attempting both didn’t have to get on airplanes between meets, although perhaps they did anyway.)
College distance coaches made a range of decisions about how to allocate their distance resources at this time last year, and the overall effect is impossible to qualify. But it seems as if comparing times between 2020—the most recent “normal” NCAA regular indoor track season, even if the championships would up nixed—and 2022 offers far more shoe-specific information than comparing the 2021 and 2022 indoor seasons.
I have no idea why Bergesen chose to focus on the shortest distance event, and in a Kevin Beck move, she also transcribed Robinson’s data about the women’s 800 meters incorrectly (the 38 should be a 34). But I took a closer look at the event, and in the past three years, these are the 25th-, 50th, and 100th-best Division I NCAA men’s and women’s 800-meter times:
2020: 1:49.09, 1:49.89, 1:51.21 / 2:06.11, 2:07.36, 2:09.02
2021: 1:49.10, 1:50.12, 1:51.49 / 2:06.00, 2:07.02, 2:10.25
2022: 1:48.05, 1:48.95, 1:50.08 / 2:04.61, 2:05.78, 2:07.69
A glance at the middle of these three rows reveals why 2021 may not be the most helpful indoor season to include, at least not on its own. The more isolated the effect of shoes on the talent pool in a given event, the more the times in that event should be expected to shift by about the same amount at each ranking position. That is, if shoes alone could be shown to account for 100 percent of the performance variation between seasons, then you’d expect the times at the 25th-place, 50th-place, and 100th-place rankings to improve by about the same amount, at least in an event as short as the 800 meters. In other words, you’d expect the whole group to improve by approximately the same amount.
Other contributors to an observed performance spike, such as a decrease or increase in the size of the group with the potential to achieve a given performance—in English, a varying number of fast college kids—might introduce differences that affect different ranking positions differently. For example, it appears likely that there were significantly fewer top-rate NCAA 800-meter runners competing in the winter of 2021 than in either 2020 or 2022. Even if the rescheduled cross-country champs didn’t draw away a lot of fast half-milers last winter, it just wasn’t a campaign with enough steam to include as complete—especially given that the 2021 indoor season will always stand as the one where “a lot” or “most” athletes already had superspikes while “a lot” or “most” did not.
So, after chucking the 2021 numbers, is there evidence of a factor that seems to affect exceptional 800-meter runners to approximately the same extent?
The 2020-to-2022 time gain at the 25th, 50th, and 100th ranking positions for men and women:
1.04, 1.06, 1.13 / 1.50, 1.58, 1.33
The corresponding 2020-to-2022 performance gain at these ranking positions, expressed as percents:
0.953, 0.965, 1.016 / 1.189, 1.241, 1.031
Allowing for the fact that this is still a very limited analysis, it seems clear that some external factor is allowing collegiate 800-meter runners to perform at a level about once percent better than before. It’s safe to say that this is almost certainly the result of better footwear and little else, although it will take a few years for the effects of pandemic-extended athletic eligibility to disappear.
Trying to determine how much the shoes help in each event continues to remain a challenge. For one thing, not every runner benefits to the same exact mathematical degree from improved shoe tech (for example, Shelby Houlihan ran American records in the 1,500 meters and 5,000 meters in old-school spikes; no word on what she’ll be in when she officially returns to the sport from time-trial land in 2025). For another, it’s possible that the effect of superspikes tends to increase with increasing race distance, meaning that a 5,000-meter runner might gain a lot more than the 1 percent or so an 800-meter runner seems to get.
But I recently saw someone—I wish I knew who so I could credit him, but he might have had a username like HarkenSludge29—propose that athletes gain about 1 second per 600 meters of race distance. I’m tempted to agree with this for now, assuming that Harkensludge29 was referring to elite athletes, who cover 600 meters of their races in close to 90 seconds. Using this loose conversion factor, a world-class male might get two-thirds of a second a lap, while a top collegiate women might gain closer to a second a lap over 10,000 meters.
If you scroll through the responses to Bergesen’s tweet, you can see that some people would rather not deal with any of the work of accounting for shoe tech, and would rather simply see the new shoes banned (Bergesen), have entirely new record lists created to account for their appearance (2:24 marathoner Kellyn Taylor), or tell journalists to take notice of the shoes and not compare times from today to those of pre-superspikes days (2:24 marathoner Kara Goucher).
There was also this from Rowan DeBues-Stafford:
comparing 2021 to 2022 doesn't illustrate that well given "super spikes" were widely available in 2021 and were even being used by some schools even in 2020
DeBues-Stafford knows what he’s talking about; his wife Gabriela had at least a prototype superspike on her feet for the 2019 World Outdoor Championships in Doha, Qatar, as did every other Nike athlete who wanted them.
I don’t think people like Bergesen are against this type of technology per se. If you asked all these same players for their thoughts five years from now, it’s likely they would be surprised at their own 2022 consternation. I suspect it’s more the mysteries than the methods that frustrate observers: If there were some agreed-upon amount of time today’s athletes could be mentally penalized in each event, pundits might not be so eager to see either superspikes banned or an entirely new post-2018-or-so record list established.
There are far greater problems with all-time lists than the drubbing shoes have recently given them. Look at the all-time women’s outdoor 800-meter list. Of the top twenty marks, in my opinion, at most one belongs to a non-doped, definitively female athlete (Athing Mu’s 1:55.04). The rest of these performances represent the work of either the betesticled or participants in drug-circuses staged in various times and places around the world, including Eastern Europe and the former Soviet Union in the 1970s and 1980s and China in the 1990s. One of the Chinese on the list is even named Dong, for Confucius’s sake.
One thing that can be affirmed without controversy is that athletes shouldn’t feel as if major figures in the sport in which they participate are rooting against them for reasons beyond their control.
What are good runners supposed to do, take rapid-onset Luddite stances and get left behind by about three steps every 90 seconds? Even runners receiving athletic scholarships?
But highly influential running figures have now been acting for several unbroken years against the interests of top athletes and everyday sloggers alike, so if today’s runners are tired of hearing how illegitimate everyone thinks their feats are, they can take solace in knowing Ethiopians are widely admired despite virtually never being drug tested and winning a lot of cool stuff, such as eight medals in the men’s and women’s 1,500 meters and 3,000 meters at the just-completed World Indoor Athletics Championships (results). Supershoes are the least of the reasons to feel that the sport’s top performance lists are imperfect and always will be in a cheating-driven sports universe.