Book excerpt: ‘How Bad Do You Want It?’

In a new book from VeloPress, author Matt Fitzgerald considers the psychology behind Greg LeMond's brilliant 1989 Tour victory.

Photo: AFP

Heading out the door? Read this article on the new Outside+ app available now on iOS devices for members! Download the app.

Editor’s note: This excerpt is republished from “How Bad Do You Want It? Mastering the Psychology of Mind over Muscle,” by Matt Fitzgerald with permission of VeloPress.

Greg LeMond woke up in a strange bed. For a second or two he knew nothing more, his mind hovering in that narrative-free state of animal consciousness that greets each of us at the threshold of wakefulness. Then it all came back to him. He was in a hotel room in Versailles, France. The date was Sunday, July 23, 1989. At 4:14 that afternoon, he would compete in the final stage of the Tour de France. It was going to be the most important race of his life.

He dressed in a yellow T-shirt and baggy blue shorts and made his way down to the ground floor, where he sat at a long table and ate a hearty breakfast of pasta, bread, cereal, eggs, and coffee with his teammates on the ADR cycling team. An hour later, they were on their bikes, just cruising, loosening up their legs for later. Overcast skies loomed above them as they pedaled away from the hotel, but by the time the ride was complete the clouds had burned off and the air temperature had risen into the low 80s. LeMond later told writer Sam Abt what he told his trainer, Otto Jácome, when he returned to his lodgings.

“My legs are good. I’m going to have a very good day.”

There was plenty of time left to kill. As the second-place rider in the general classification (GC), or overall race standings, LeMond would start the conclusive 24.5km individual time trial next to last among the Tour’s 134 surviving competitors, two minutes before Frenchman Laurent Fignon, the race leader. A two-time winner of the Tour de France, Fignon stood merely 50 seconds ahead of LeMond after 20 stages and more than 2,000 miles of riding. LeMond was the stronger time trialist, but he would have to make up an improbable two seconds per kilometer between Versailles and Paris to overtake Fignon in the GC and claim his own second Tour de France victory. Whichever way it went, it promised to be the closest finish in the event’s 76-year history.

A 24.5km cycling time trial is an exercise in pacing. So are all races that last longer than 30 seconds. In races that last less than 30 seconds, competitors go all-out, pedaling, striding, or stroking at absolute maximum intensity from start to finish. They hold nothing back and utilize their full physical capacity. In races that last longer than 30 seconds, competitors do hold back. They pedal, stride, or stroke at less than maximum intensity at all points of the race except perhaps the very end. Instead of going all-out, they maintain the highest intensity they feel capable of sustaining through the full race distance.

Why 30 seconds? Because humans cannot sustain maximum intensity exercise longer than about 30 seconds without exceeding the highest level of perceived effort they can tolerate. Athletes are conscious of their effort in shorter races, of course, but because they know their suffering will end quickly they do not use this perception to control their pace, which is constrained only by their physical capacity. But when an athlete starts a race that he knows will last longer than 30 seconds, he holds back just enough that his perceived effort limit is not reached until he is at the finish line. That is the art of pacing.

What happens when an athlete tries to sustain a maximum intensity of exercise longer than 30 seconds? Anna Wittekind of the University of Essex answered this question in a 2009 study published in the British Journal of Sports Medicine. Nine subjects were asked to ride stationary bikes outfitted with power meters as hard as they could for 5 seconds, 15 seconds, 30 seconds, and 45 seconds on separate occasions. When she reviewed the results, Wittekind found that the subjects had generated slightly less power during the first 15 seconds of the 45-second test than they had in the 15-second test. In other words, they had not pedaled as hard as they could at the start of the longest test ride, even though they had been instructed to do so. Instead, they had unconsciously paced themselves.

Wittekind speculated that, on the basis of past experience, the subjects recognized that they could not sustain a true maximal effort for 45 seconds without exceeding their maximum tolerance for perceived effort, so they held back just a little without even realizing it. These results suggest that the limit of maximum perceived effort tolerance is so impenetrable that athletes are not psychologically capable of even trying to sustain a maximum exercise intensity longer than approximately 30 seconds.

The fact that pacing is required to maximize performance in all races lasting longer than half a minute has some interesting implications. A sprinter finishes every race knowing he went as fast as he could (technical errors notwithstanding). Longer races are different. Because it is necessary to hold back to some degree at almost every point in these races, it is impossible for the athlete to know upon finishing whether he might have gone faster — if only by a second or two — if he’d held back just a bit less somewhere along the way.

Many automobiles have a “range” feature that displays the number of miles the vehicle can travel before it runs out of gasoline. This number is not open to interpretation. If the range display says 29 miles, you’d better find a gas station within the next 29 miles. The mechanism of regulatory anticipation that athletes use to control their pace in races is different. It’s not a number but a feeling, and like all feelings it is open to interpretation. One of the most important and valuable coping skills in endurance sports is the ability to interpret the perceptions that influence pacing decisions in a performance-maximizing way. As an endurance athlete, you want to get better and better at reading these perceptions in such a way that your internal pacing mechanism functions more and more like an automobile’s range display. You want to be as correct as you can possibly be when determining the swiftest pace you can sustain to the finish line without exceeding your perceived effort tolerance.

Setting and pursuing time-based race goals is very helpful in the process of calibrating anticipatory regulation. This practice enables athletes to interpret their effort perceptions in a more performance-enhancing way by transforming the racing experience from an effort to go as fast as possible into an effort to go faster than ever before. Validation of this approach comes from a 1997 study done by researchers at Israel’s Ben-Gurion University and published in the Journal of Sports Science. High school students were subjected to a test of muscular endurance and then spent eight weeks training to increase their time to exhaustion. Some of the students were given a nonquantitative goal to “do their best.” Others were given a quantitative goal to better their performance in the initial test by a certain percentage. Even though all of the students did the same training, those who pursued quantitative goals improved their performance significantly more when the muscular endurance test was repeated after eight weeks.

More recently, a team of researchers led by Eric Allen of the University of California found that finish times in marathons tend to cluster near the round numbers (such as 4:30 and 4:00) that runners typically pursue as goals. This pattern would have carried little significance if Allen and his colleagues had not also noted that those runners who end up closest to these round numbers at the finish line slow down less than other runners in the final miles of a marathon — evidence that the pursuit of these round-number goals enhances performance. Regardless of how an athlete chooses to train, her training will yield greater improvement in race times if improving race times is the explicit goal of the training process.

Paying attention to the clock reduces the uncertainty associated with reaching beyond past limits and in this way facilitates effective pacing. While it may be impossible for an athlete who completes a race of a certain distance to know if he could have tried harder, it’s relatively easy for an athlete who completes a race of a certain distance in a certain time to aim to cover the next race of the same distance a second or two faster than he did the last time.

According to Samuele Marcora’s psychobiological model of endurance performance, the amount of effort that an athlete puts into a race is influenced by her perception of the attainability of her goal, a concept borrowed from Jack Brehm’s theory of motivational intensity. If the goal seems to fall out of reach at any point during the race, the athlete is likely to back off her effort. If the goal seems attainable, but only with increased effort, the athlete is likely to increase her effort, provided she’s not already at her limit. By keeping track of, and aiming to improve, personal best times for specific race distances, athletes can exploit this phenomenon to try harder than they would otherwise be able to. The goal of improving your time for a certain distance by 1 measly second almost always seems attainable. And if that goal is attainable, then the very slightly greater level of perceived effort that an athlete must endure to achieve it is likely to seem more endurable than it would seem if the athlete were going entirely by feel. It’s not the time goal itself that enhances performance but the effect that the goal has on how the athlete interprets her perception of effort.

Setting time-based goals that stretch you just beyond past limits is like setting a flag next to a bed of hot coals to mark the furthest point reached in your best fire walk. That flag says to you, “This is possible, and you know it. So why wouldn’t it be possible for you to make it just one step farther the next time?”

A real-world example of this process of using time-based goals to recalibrate perceived effort in a performance-enhancing way is South African runner Elana Meyer’s career progression at the half-marathon distance. In 1980, when she was 13 years old, Meyer took her first shot at 13.1 miles, winning the Foot of Africa half-marathon in a mind-boggling time of 1:27:10. Nine years later, Meyer made her professional debut at the same distance, running 1:09:26 in Durban. In 1991, she smashed the half-marathon world record in London, clocking 1:07:59. Between 1997 and 1999, Meyer broke the record thrice more, running 1:07:36, 1:07:29, and finally 1:06:44 in Tokyo at age 32.

Obviously, Meyer’s development as an athlete was responsible for much of this improvement. But her pursuit of time goals also played a role. It is interesting to note that her margins of improvement tended to get smaller as her career advanced. Her big leaps from 1:27 to 1:09 and from 1:09 to 1:07 were undoubtedly fueled principally by gains in fitness. Meyer probably wasn’t even thinking about her first half-marathon when she made her pro debut, so much stronger was she by then. But her last two world records were set on familiar courses on which she had already posted fast times, and in each of these cases she set out deliberately to run faster than ever before. It’s likely that Meyer was not any fitter at the 1999 Tokyo Half Marathon, where she ran 1:06:44, than she had been a year earlier in Kyoto, where she ran 1:07:29, but she had the crucial advantage of having run 1:07:29 already.

But wait: If Meyer was just as fit (not to mention a year younger) when she ran the slower time, then can it not be said that timekeeping held her back in the 1998 Kyoto Half Marathon, even as it pulled her beyond the world record of 1:07:36 she had set on the same course in 1997? There is indeed evidence that the influence of clock watching on endurance performance is two-sided. The same time goal that enhances performance when it is perceived as a target constrains performance when it is perceived as a limit.

The potential for time standards to become performance limiters is most apparent at the elite level of endurance sports. There have been many noteworthy cases in which a performance breakthrough by one athlete triggered a widespread revolution in performance and thereby revealed that previous standards had been holding the sport back. Between 1994 and 2008, for example, the women’s world record for triathlon’s Ironman distance was stuck at 8:50:53. Only seven women recorded times under 9 hours in that 14-year span. When Yvonne van Vlerken finally lowered the Ironman world record to 8:45:48 in July 2008, the floodgates were opened. Six other women dipped under the 9-hour barrier in the next few months. Van Vlerken’s mark lasted only one year, as did the subsequent record. By the end of the 2011 season, the Ironman world record for women stood at 8:18:13, and sub-8:50 performances had become commonplace. Was the new generation of female triathletes that much more talented than the previous one? No. These women just weren’t held back by a tendency to regard the time of 8:50:53 as an unsurpassable human limit.

In consideration of the two-sided nature of time’s effect on endurance performance, it is tempting to ask what sort of time goal would have the best possible effect on performance. Such a goal would need to seem reachable, but barely so. (Indeed, in the Ben-Gurion University study I mentioned above, students given a “difficult/realistic” goal improved more than those given either an “easy” goal or an “improbable/unattainable” goal.) This ideal goal would also need to be sufficiently well defined to pull the athlete beyond past limits, yet somehow vague enough that it did not place an artificial ceiling on the athlete’s performance.

LeMond’s situation at the start of stage 21 of the 1989 Tour de France met these requirements perfectly. LeMond had to beat Laurent Fignon’s time in the 24.5km time trial by 50 seconds. But Fignon would start behind him, so LeMond could not approach the race with a specific time in mind, such as the 27:30 clocking that Thierry Marie posted early in the day, which stood as the best time in the field when LeMond started his ride. Instead, LeMond knew only that he had to ride 50 seconds faster than the best time Fignon—one of the world’s best time trialists besides LeMond himself — could conceivably achieve on his best day.

LeMond told reporters before the race that he believed the task facing him lay at the very outer limits of the achievable. He was not certain that he could pull it off even if he gave more than he had ever given before on a day when he had more to give than ever before. Nor was he certain that he couldn’t. It is hard to imagine a goal construct that would have elicited a better performance from LeMond in the most important race of his life.

After lunch, LeMond checked out of the hotel and made his way toward the Palace of Versailles, a Vatican-like architectural colossus in front of which a comparatively flimsy temporary starting platform had been erected under a white canopy. A massive crowd had gathered there to witness the showdown between the two men at the top of the overall. Behind the start line was a small warm-up area. Within its narrow confines a handful of riders traced tight loops. LeMond joined them and soon met Fignon head on. LeMond averted his gaze. Despite this demurral, Fignon thought the American looked relaxed. In fact, LeMond was terrified, his stomach knotted with dread of the suffering he was about to inflict on himself.

At 4:12, Pedro Delgado, who stood 1:38 behind LeMond in the GC, rolled off the starting ramp and accelerated down the broad Avenue de Paris. It was now LeMond’s turn to mount the platform. Television cameras rolled as a silver-haired race official with black-framed glasses held LeMond’s bright red Bottecchia time trial bike upright and a countdown was intoned over loudspeakers.

“Cinq … quatre … troix … deux … un … Allez!”

LeMond stood on the pedals and began a hard windup, his feet churning like the steel wheels of an accelerating locomotive. When he hit 100 revolutions per minute, he dropped his butt onto the saddle and settled his forearms into the aero bars. A pair of police motorcycles guided him down the runway-wide boulevard as a flotilla of vehicles, including a white Peugeot containing ADR team manager José de Cauwer, followed behind.

LeMond’s plan was simple: to ride just a bit faster than he ever had, holding back a little less than he had ever dared in similar circumstances. LeMond lowered his head and rode with his eyes cast straight downward, as though indifferent to where he was going, looking up only briefly every few seconds to check his line. His meaty quadriceps billowed with every downstroke.

LeMond was already more than a mile down the road toward the Parisian suburb of Viroflay when Fignon set off behind him. With his granny glasses and blond ponytail, he looked more like a high school drama teacher than a professional cyclist as he sprinted away from the starting gate. Fignon had been seen fiddling around with a set of triathlon bars earlier in the day, but he’d elected to leave them behind. His bike did have the advantage of being outfitted with two aerodynamic disk wheels, however, whereas LeMond, expecting more crosswinds than he actually would encounter on the course, had gone with spokes in the front.

Fignon felt strong and confident. He zipped through Viroflay and approached Chaville, the crowds thinning as he went. After he passed five kilometers, his team manager, Cyrille Guimard, shouted from the trailing car, his words captured by a nearby cameraman’s mic.

“Six secondes!” he called out. “Vous avez perdu six secondes!”

He had already lost 6 seconds to LeMond. Fignon turned his head and stared incredulously at Guimard. LeMond was not yet gaining the 2 seconds per kilometer he needed to overtake the Frenchman, but given how well Fignon himself was riding, he couldn’t believe the American was going that much faster.

Up the road, LeMond received the same news from José de Cauwer. Before the race, LeMond had asked Cauwer not to supply any such information, and he tersely reminded his manager of that wish now. For the remainder of the time trial, his mind would be focused entirely on the image of creating distance between himself and the man behind him, and on the only number that mattered: 50 seconds.

Passing through Sèvres, on the west bank of the River Seine, LeMond came to an overpass. He moved his hands to the outer bars and pedaled from a standing position to avoid losing speed as he climbed. If the policeman on a motorcycle cruising close behind him had checked his speed gauge at this moment, he would have seen the needle fixed at 54 kilometers per hour.

Minutes later, LeMond crossed the Pont de Sèvres, a bridge over the River Seine, at the far end of which he made a sharp right turn onto the Quai Georges Gorse, carving the corner with such bold precision that his right shoulder came within centimeters of clipping spectators leaning against a barrier on the inside of the turn.

At 11.5 kilometers, LeMond passed an official time check. His split of 12:08 was the fastest of the day by 20 seconds. Fignon reached the same point 2 minutes and 21 seconds later, having now lost 21 seconds to LeMond since the start. If Fignon continued to lose time at the same rate, he would complete the time trial 45 seconds slower than LeMond and would win the Tour de France by five seconds.

The Quai was as flat and straight as a drag strip. LeMond took full advantage, settling into a chugging rhythm that nudged his speed even higher. The drivers of any trailing cars with manual transmission would have been forced to shift into fourth gear to keep up.

As LeMond approached 14 kilometers, the Eiffel Tower rose into view ahead. Fignon’s deficit had risen to 24 seconds. LeMond still was not gaining time quite fast enough, but as hard as he was pushing himself, he still felt strong, whereas Fignon’s shoulders had begun to rock, a telltale sign of encroaching weariness. A clear difference in the relative speeds of the two men became apparent to cycling fans watching the battle on television at home. Every 10 or 15 seconds, the coverage jumped from Fignon to LeMond, and when it did, the passing scenery accelerated noticeably.

The last part of the course skirted the famous Jardin des Tuileries — Paris’s Central Park — and dumped riders onto the Champs-Élysées for the homestretch to the finish line. Tens of thousands of spectators lining the streets there erupted when LeMond came into view. (His French surname, French language skills, and all-American charm had won him many admirers in the Tour’s host nation.) He passed under a banner marking four kilometers to the finish line. His advantage was now 35 seconds. LeMond had stolen exactly two seconds per kilometer from Fignon over the last 6.5 kilometers; he would have to nearly double that rate of separation on the Champs-Élysées to win the Tour.

LeMond’s last best chance to gain that separation lay just ahead of him, at three kilometers to go, where a false flat rose gently from the Place de la Concorde to the Arc de Triomphe. It wasn’t much of a hill by Tour standards, but to an exhausted rider — as Fignon was quickly becoming — it would feel like a Pyrenean switchback. LeMond attacked it hard, telling himself that his career depended on it. As he neared the top, his torso began to pump up and down like an oil horse. Any consideration of good form had gone out the window — all that mattered now was effort at any cost.

At the Arc de Triomphe, LeMond made a hairpin right turn and entered the final straight to the finish line. Moving down the same false flat he had just ascended, he hit 40 mph, approaching the motor vehicle speed limit on the Champs-Élysées. He passed under the 1-km banner. Over the race radio came word that LeMond still needed 10 seconds.

Ahead on the road, LeMond saw the rocking posterior of Pedro Delgado, who had started two minutes before him. LeMond felt a magnetic pull, and he used it to raise his effort level one more excruciating notch for the final drive to the finish line. He crossed at 26:57, beating the previous best time of the day by 33 seconds. LeMond hung his head like a recipient of bad news as he coasted to a stop. A moment earlier his legs had felt as though they were going to explode. Now they suddenly felt capable of going another 10 miles. Had he done enough?

The waiting began. LeMond dismounted and turned back toward the racecourse and the finish line clock, understanding that if it displayed the number 27:47 before Fignon finished, he had won the Tour de France. The anticipation was unbearable. When Fignon came into sight, LeMond reflexively shaded his eyes and looked away — but only briefly.

Fignon was shattered with fatigue, no longer able to hold a straight line and nearly drifting into a barrier of scaffolding at the outer edge of the 70-meter-wide road in his flailing efforts to drive his machine toward the line. The seconds ticked by with surreal slowness. But the magic number finally appeared, and when it did, Fignon was still 100 meters from the finish. He stopped the clock at 27:55. LeMond had won the Tour by eight seconds.

LeMond’s average speed for the 24.5km time trial was 33.89 mph — an all-time record for Tour de France time trials, by a long shot. And to this day, no Tour rider has ridden faster than LeMond except in shorter time trials undertaken on fresh legs on the first day of the race instead of the last.

It would seem that, in the right circumstances, an old-fashioned stopwatch — properly used — can affect endurance performance more powerfully than either the finest equipment or the most potent chemicals — not to mention lift an athlete to his finest hour even after his best days are behind him.

Trending on Velo

An American in France

What’s it like to be an American cyclist living in France? Watch to get professional road cyclist Joe Dombrowski’s view.