The Psychology of Racing Versus Pacing
Running alone against the clock is very different from trying to beat other runners, but untangling how our minds process the challenge is “like knitting with spaghetti”
Receive $50 off an eligible $100 purchase at the Outside Shop, where you’ll find a selection of brand-name products curated by our gear editors, when you sign up for Outside+ today.
A year ago, when the idea of a “virtual race” seemed like a novel concept rather than a sick joke, I wrote about a study that explored the psychological differences between solo time trials and head-to-head races. A key observation: effort (how easy or hard it felt) was the same in both situations, but affect (how good or bad it felt) was very different. The power of running with others is that it can make a hard effort feel good, or at least less bad.
Now the same research team, led by Everton do Carmo of Senac University Center in Brazil, has a new study in the European Journal of Sport Science that digs further into the topic—and specifically into the question of goals. Anyone who has watched the cat-and-mouse tactical games in middle-distance track races at the Olympics knows that trying to win and trying to run fast produce very different styles of race. And there’s also a big difference between racing a stronger opponent and racing a weaker one. As you add more and more variables into the mix, the psychology of pacing gets very complicated—and interesting patterns emerge.
The new study put 13 male cyclists through a series of 10K races in a virtual reality setup over the course of a few weeks. They did two solo time trials around a 250-meter virtual velodrome, and two head-to-head races against a virtual opponent. In one case, the opponent was programmed to go exactly six percent faster than the subject’s best solo time trial; in the other case, they went exactly three percent slower. In addition to measuring performance, the researchers quizzed the subjects once every kilometer about a set of psychological variables: perceived effort, affect, and self-efficacy, which is essentially the degree to which you believe you can successfully meet a performance goal.
The top-line result is a bit befuddling: the subjects recorded pretty much identical times, on average, in all three conditions. This conflicts with the study I wrote about last year, in which runners went faster with competition than they did alone. It also conflicts with numerous other studies, and with the lived experience of the vast majority of endurance athletes (though not everyone, as I heard last time I wrote about this topic!). The reason is very likely that the performance gaps were too big: the fast opponent was impossible to beat, and the slow opponent was no challenge. There’s some previous evidence for this: several studies have found that racing against a virtual self going two percent faster improves performance, but racing against a five-percent-faster opponent doesn’t.
Still, despite the similar finishing times, there were some telling differences in how they got there. For starters, while the overall pacing pattern (fast start, slow middle, fast finish) was consistent, racing against an opponent led to a faster start. Here’s what the pacing pattern looked liked for the solo time trial (TT), racing against the slower opponent (SLOW), and racing against the faster opponent (FAST):
Very roughly, it looks like the head-to-head racers boosted their power output by about six percent (~330 vs. 310 watts) in the first kilometer. That makes sense when you’re riding against an opponent who is (unbeknownst to you) riding six percent faster than your usual pace—but it’s surprising that the same thing occurs when riding against the slower opponent. Rather than a rational adjustment of speed to match the opponent, this looks more like a knee-jerk response to the challenge of trying to beat somebody: competitive juices trumping the usual time-based pacing instincts.
That brings to mind the Letsrun message board report that a Youngstown State runner named Chase Easterling ran the first mile of the NCAA cross-country championships earlier this month in a blistering 4:38—but was in last place among the 255 entrants at that point in the race. It’s hard to imagine that this pace was optimal for more than a handful of the runners in the field. Of course, you have to weigh that against the reality that positioning matters when you’re cramming 255 people into a series of narrow paths and trails. Pacing decisions don’t occur in a vacuum—but even in the sterile confines of the lab, the prospect of racing against someone else seems to prod us to sprint off the start line.
There’s one other interesting detail in that pacing data above. Look at the tenth and final kilometer, on the far right. As expected, the subjects accelerate as the finish approaches. In the head-to-head races, the finishing sprint is much less pronounced, perhaps because they’re paying for their aggressive start. In the race against the slower opponent, where the primary goal was to win, it might be that no finishing sprint was needed because the subjects were already well ahead. But in the race against the fast opponent, the final kilometer is actually slower than the previous one. Is this a sign that starting fast and desperately trying to keep up with a faster opponent pushed the subjects to their absolute limits, leaving nothing for a finishing sprint?
Not quite. Take a look at the data on rating of perceived exertion (RPE, on a scale of 6 to 20), which climbs steadily from a relatively light initial effort to a near-maximal finish:
In the final three kilometers, you can see the level of effort when racing against the faster opponent starts to tail off. The difference isn’t statistically significant, but it appears that by the last few kilometers of the race it becomes clear that they’re not going to catch up with their unexpectedly strong opponent. They know they’re going to lose, and the slightly lower effort they’re willing to put out reflects that realization. That’s why the power output drops in the final kilometer.
You might think they’re slacking off near the end because they’re not having fun anymore. In the study I wrote about last year, affect—the sense of positive or negative feelings—declined steadily when racing alone but stayed stable when racing in a group. In this case, though, affect declined at a similar rate in all three groups. Running or cycling in a pack may be pleasant, but getting smoked in a one-on-one duel, even by a virtual opponent, doesn’t seem to elicit the same happy feelings. The biggest drop in affect was in the group racing against a faster opponent, but the differences compared to racing alone or against a slower opponent weren’t huge: affect wasn’t the difference-maker.
There’s one last variable: self-efficacy. How confident are you in your ability to complete the task and achieve your goal? At the start of the race, everyone feels pretty good about their chances. But once you start racing someone who’s six percent faster than your own previous best, it’s hard to keep your chin up. Here’s the self-efficacy data:
It’s a bit tricky to sort out chicken and egg here. High self-efficacy is supposed to be beneficial for performance; but in this case, the steadily declining self-efficacy of the fast-opponent group just seems like a rational acknowledgement of reality. At some point, insisting “Yes, I can beat that guy” shifts from optimism to delusion.
The takeaways here aren’t straightforward—which, perhaps, is the point. In past articles, I’ve highlighted the role of perceived effort as the “master switch” that controls endurance performance and dictates what pace you can sustain. That may be true in the lab, where other variables are carefully controlled. But in the real world, your pacing will be affected by the situation, the presence and actions of other people, and the goals you’ve set for yourself that day.
I asked University of Worcester researcher Andy Renfree, a co-author of the new study, what he took from it. “My personal feeling is that everything follows from goal setting,” he replied, “but untangling the relationships between RPE [i.e. effort], affect, and self-efficacy is very complicated.” In the words of one of his colleagues, he added, “it’s like knitting with spaghetti.” That’s undoubtedly true—but I do think we can pull a few useful strands out of studies like this one. Mass participation races are somewhere on the horizon, and when they arrive, try not to show your enthusiasm by sprinting the first mile in 4:38. Aim to beat someone who is two percent faster than you. And, if possible, enjoy it.
For more Sweat Science, join me on Twitter and Facebook, sign up for the email newsletter, and check out my book Endure: Mind, Body, and the Curiously Elastic Limits of Human Performance.