False starts are a bane of athletics.
In sprint races which can be won or lost by hundredths of a second, the need to start quickly is important, but the penalty for starting too soon is draconian: in modern sports subject to timing by sophisticated timing devices, one mistake and you’re out.
False starts are seared into the minds of Australia’s sports-loving public due to the outrage and misunderstandings they generate. Superstar swimmer Ian Thorpe literally fell off his blocks! And in the 2010 Commonwealth Games in Delhi, Australian sprinter Sally Pearson left the stadium in tears after being charged with a false start in the 100m.
More recently, American 110m hurdle hopeful Devon Allen was summarily tossed out of the final of the 2022 World Championships in Oregon in the USA, for being one millisecond too quick off the blocks, less than one percent the duration of the proverbial blink of an eye.
Until a few decades ago, false starts were called by starting-line officials trained to watch for runners who moved before the sound of the gun. Now for track athletes they’re determined by sensors in the starting blocks designed to measure if runners are pushing off too soon.
The assumption is that nobody can react faster than 100ms, so anyone who starts to push off before then must have anticipated the gun, however slightly, even if no human eye could have caught it. Allen was disqualified because he started to push off at 99ms.
The science behind this stems from a 1990 Finnish study of eight elite male sprinters, whose average reaction times were 121ms. The 100ms threshold was picked partly to account for variability across athletes, and partly because it was a nice, round number.
The basic idea is sound. “There is definitely a delay between the gun and a human being responding, and it’s a significant delay,” says Andrew Harrison, a professor of sports biomechanics at the University of Limerick, Ireland.
But is 99ms truly impossible?
Reaction time is based on a number of factors, starting with the speed of sound. You can’t react until you hear the gun, and fast as it is, sound progresses at only about 3ms per meter.
That means with a traditional starter standing on the infield, a runner in lane eight has about a 20ms disadvantage over the person in lane one. That doesn’t sound like much, but at world-class sprint speeds, it’s 20 centimeters at the finish potentially the difference between a medal and off the podium.
To correct this, the starting “gun” is now an electronic bang broadcast simultaneously from speakers behind each athlete. Still, you can’t put the speakers so close that there is no speed-of-sound delay. Matthew Pain, a biomechanics researcher at Loughborough University in the UK, puts it at 3 to 5ms.
After this, however, things start getting sticky because there is a difference between what Pain calls “voluntary reaction time” and the “startle reflex” which bypasses the thinking part of the brain and goes straight into the spinal cord. It’s not clear which pathway any given athlete uses, but the startle reflex is fast enough, Pain says, that in laboratory studies, nerve signals can be detected reaching all the way to the calves as quickly as 60ms after the sound is heard.
Not that muscles actually start doing anything that quickly: it takes at least another ten milliseconds for them to respond and start contracting. After that, it might take another five to 10ms to produce detectable force.
Still, if you add it all up, it’s a lot faster than 100ms: more like 78 to 85ms, Pain says.
Bas van Hooren, a sports scientist at Maastricht University, The Netherlands, agrees. “When we sum all delays,” he writes on his blog, “we arrive at a minimum response time of approximately 84ms between the start signal and the first application of force to the starting blocks.”
Starting blocks, however, aren’t designed to detect the first hint of motion. If they were, they’d respond to minute shifts in the athlete’s balance, and pretty much everyone would be disqualified, because nobody can hold perfectly still. — Pain says.
To do that, the blocks are set to not sound the alarm unless the runner has, within 100ms, made a substantial move to push off against them. Depending on how you set the threshold, Pain says, that could add at least another 25ms onto lab-measured reaction times, bumping the fastest-possible block-measured response time up to comfortably above 100ms.
“You’re starting to talk about 110 or 120ms,” he says.
That said, lab tests have shown that the variability in how fast people can increase pressure on the blocks “is huge”—not based not so much on how fast they react to the gun, but on how fast their muscles can ramp up the pressure.
“You can have a seven-fold or eight-fold difference,” he says.
That means someone whose muscles build power quickly will hit the blocks’ force threshold much sooner than someone whose muscles are slower to react. Also problematic is that there is no rule dictating how to set the threshold, and that the thresholds used by different block manufacturers are closely guarded secrets.
Starting block readings are retained and available for later analysis
It appears the World Championships may have used unusually sensitive settings that may well have cost Allen his shot at a medal.
In Eugene, Oregon, where the championships were held, Pain says everything including the stadium and the timing system had been rebuilt.
“Seiko made a big deal about their redoing the Eugene stadium,” he says. “There was a lot of advertising about how they got their new system in there.”
Normally, block-measured reaction times coming even close to the 100-milliseond threshold are rare.
In the previous five or six World Championships, Pain says, there’d been five or so who’d managed to react faster than 115ms. In Eugene, there were 25. Devon Allen himself had a (legal) reaction time of 101 milliseconds in a heat, before he was disqualified for being two milliseconds faster in the final.
“You can’t say the threshold was changed because they won’t tell you,” Pain says. “They will tell you it’s been calibrated, it’s fine. No, there was something different.”
It wasn’t just at the World Championships, he adds, because the U.S. Olympic Trials held the previous year, in the same stadium and with the same timing equipment, also had a large number of false starts.
And a lot of unusually quick legal starts.
Supporting the idea that this might have been the result of changes in one stadium’s timing equipment, he adds, is the fact that the Commonwealth Games, in Birmingham, UK, had no problem with false starts and a more normal range of block-measured reaction times, as did the recently completed European Championships in Munich, Germany.
“Reaction force under the hands always precedes [that measured by] the blocks.”
Professor Andrew Harrison, University of Limerick
What to do about this is an open question.
“I don’t mind 100ms,” Pain says, “but you have to take responsibility for your [measurement] system being part of that.”
Harrison would like to address the issue by changing more than the parameters used by the measurement system. The feet aren’t the only part of the body in contact with the ground at the start.
Why not put pressure plates under the fingers? “They’ll respond first,” he says. “Reaction force under the hands always precedes [that measured by] the blocks.”