Lazarus Lake on Endurance, Uncertainty, and Reaching One’s Potential

Lazarus Lake — Gary Cantrell, designer of the Barkley Marathons and Big’s Backyard Ultra — talks with Tyler Cowen about why he builds races that defeat almost everyone who attempts them, and what four decades of running them have taught him about how far people can actually go. The conversation is less about Barkley folklore than about a design problem: how to hold an event at the edge of the barely possible, and what role deliberately engineered uncertainty plays in getting entrants there.

Key ideas

1. The race is built to test self-reliance, not to be conquered. Lake frames these events around total dependence on oneself, with modern conveniences stripped away, tested against a roughly 1 percent finish rate — ‘you have selected the best of the best, and out of those, only the 1 percent make it.’ He attributes the heavy STEM representation among ultrarunners to the events being ‘a problem-solving exercise’ — map-and-compass navigation, resource management, an 80-degree temperature swing on a single loop — rather than a pure fitness test.
2. The field is tuned to hit a lap target, and the target is set by the wilderness, not by difficulty. The course has stayed largely the same since 1986; what Lake tunes is who gets in. Entrants are screened by essay rather than credentials, then sorted into ability bands so the total number of laps run across the field lands near about 190. His stated reason is environmental: the race’s impact on the wilderness area is, he says, ‘really dependent not on the number of people but the number of laps’. The selection process is how he hits that figure, and he reports hitting it reliably — a year with three finishers, unusually many, still produced fewer total laps than the year before, which had none.
3. Uncertainty is engineered in on purpose. The course is not disclosed until the day before the race; the start time is signalled only by a conch shell blown an hour beforehand, within a window entrants know only as ‘between midnight and noon’; the following year’s date is known only to Lake and his successor. His reasoning: ‘uncertainty is the most difficult thing for an athlete to deal with’ — so removing every fixed point an entrant could plan around is itself part of the test.
4. He asserts, more than proves, that people routinely underestimate what they can do. His central claim — ‘each time you step up a level, you find out that you can do a lot more than you thought you could do’ — he extends, when pressed, to ‘all’ human endeavour. He offers no evidence for the generalisation beyond his own decades of watching entrants; it functions as a working belief, not a measured finding.
5. On the outer limit of human performance, he professes not to know. Asked how close ultrarunning is to the fundamental limits of human performance, Lake’s answer is ‘I don’t know.’ The closest thing to a settled position is that ‘there’s this ability of the mind to drive the body beyond what seems to be physically possible’ — treated as an open question the sport exists to probe, not an answer he holds.
6. He rejects the standard self-talk prescription in favour of doubt. Pressed on research showing affirming self-talk helps performance, Lake disagrees for himself: he prefers ‘a little bit of doubt’, reasoning that certainty removes the point of the effort — when people told him ‘you’ve got this’ near the end of a coast-to-coast walk, his response was to ask why he should keep going if he already knew he’d succeed.

Content

Why design a race almost nobody finishes

Cowen opens by asking what running 100 miles teaches that 26 miles does not; Lake’s answer sets the frame for the whole conversation — scale changes the lesson qualitatively, not just quantitatively: ‘it’s not really an extension of the same thing.’ Runners come away, in his account, having learned they can do more than they believed, and he generalises this past running to ‘most human spheres of endeavor.’

Asked why STEM professionals are so heavily represented among ultrarunners, Lake locates the appeal in the events being ‘a problem-solving exercise’ rather than a pure fitness contest: competitors must navigate by map and compass, manage water and clothing without aid stations, and cope with an 80-degree swing in temperature on a single loop through Frozen Head State Park’s steep, undergrowth-choked hills. The reward he describes is not the finish line but being ‘totally dependent on yourself’ — set against a roughly 1 percent finish rate that lets each entrant test whether they are ‘1 percent of the 1 percent.‘

Selecting the field, and the constraint that shapes it

The course has stayed largely the same since 1986, so what Lake adjusts each year is who gets in. Applications are screened by essay rather than credentials or references: ‘even if someone’s not particularly eloquent, if they write their essay, they reveal themselves.’ He is filtering for entrants who will get something from the experience against those who want only to say they entered, or to withdraw before it hurts.

The reason the composition matters is not, on his account, difficulty. It is the wilderness. Lake does not want the race to mark the wilderness area it runs through, and that impact, he says, is ‘really dependent not on the number of people but the number of laps’ — so entrants are sorted into ability bands chosen to bring the field’s total laps out near 190. It is an environmental target, hit by picking who runs rather than by changing the course. Within each band the draw is random: he uses Excel’s random number generator.

He reports hitting the figure reliably, and the evidence he offers cuts against the intuitive link between finishers and effort: ‘Even this year, when we had three finishers, which is really unusual, we still had just about the same number of total laps that we always have. It was fewer, actually, total laps than we had last year with zero finishers.’ A year almost nobody finishes and a year three people do are, in aggregate, roughly the same amount of running. What Lake does not claim — and the transcript does not support — is that this selection is a dial for tuning how hard the race is; he never describes it that way.

Uncertainty as the real design material

The title’s second theme gets its most direct treatment near the end of the conversation, prompted by the ritual of the conch shell that signals the race is an hour from starting. Lake states the underlying principle plainly: ‘uncertainty is the most difficult thing for an athlete to deal with,’ and the Barkley is built to maximise it rather than remove it. The course is not revealed until the day before the race. The start time is disclosed to no one — entrants know only that it will fall ‘between midnight and noon’ — so the hours before a race that might start at dawn produce sleepless nights spent listening for the conch. The following year’s date is known only to Lake and his successor, Carl Laniak, until entrants are told.

This is a different use of uncertainty from a competitor weighing an unknown probability before acting: Lake is not the one deciding under uncertainty, he is manufacturing it for others to endure. The tracker regulations European organisers face, in his telling, blunt this design lever — even though a tracker cannot tell a lost runner where they are, knowing that someone else can locks in a fallback the Barkley deliberately withholds. ‘Nobody knows where they are’ is the condition the race is built to preserve.

What he claims about human potential — and what he does not

The title’s other claim, ‘reaching one’s potential,’ rests on two things Lake asserts without evidence beyond his own observation: that people routinely underestimate what they can do, and that this holds across domains, not just running. Both are stated as settled personal conviction — ‘I think it would be true in all of them’ — rather than argued from data.

On the harder question — where the actual ceiling on human performance sits — he does not claim to know. Asked directly how close ultrarunning is to the fundamental limits of human performance, his answer is ‘I don’t know,’ followed by a description of the sport’s purpose rather than a finding: people ‘can learn and adapt to become better,’ and ‘there’s this ability of the mind to drive the body beyond what seems to be physically possible. Who knows where the real limit lies?’ Even on a fact he might be expected to hold with certainty — whether the Barkley is finishable at 50 — he first doubts it, then concedes David Horton did it on his eleventh attempt, and settles on ‘we like to do the impossible’ rather than revising his model of what is possible.

Where Cowen presses and Lake pushes back

Two moments read as genuine disagreement rather than deflection. Cowen raises research on affirming self-talk; Lake does not adopt it, saying he prefers to ‘approach things with a little bit of doubt’ because certainty removes the reason to keep going — when people told him ‘you’ve got this’ near the end of his walk across America, his reaction was to ask why he should endure the discomfort if the outcome was already known.

Elsewhere Lake declines to answer rather than push back. Asked to name the individual sports performance that impresses him most, he says he would need to ‘sit and really think about that a long time’ and that different running disciplines alone are not easily comparable, let alone different sports. On his own rising fame over the past decade, he resists Cowen’s frame that it reflects achievement, calling it ‘a combination of circumstances’ rather than something he earned. And asked about the enduring media link between his races and escaped convict James Earl Ray, who evaded capture for days in the same mountains, Lake calls the connection incidental and says drawing it risks ‘belittling his escape attempt more than anything else’ — declining an opening to lean into the story’s mythology.

See also

• Lazarus Lake — speaker; race designer and former ultrarunner
• Tyler Cowen — host
• Vishy Anand — fellow Conversations with Tyler guest on sustaining elite performance past the point most competitors fade
• Vishy Anand on Staying in the Game — companion episode on rebuilding a practice to stay competitive with age, in a different demanding discipline
• Knightian Uncertainty — nearest wiki neighbour on uncertainty, though the mechanism differs: that concept names the irreducible epistemic uncertainty a decision-maker lives inside, where Lake’s is uncertainty a designer deliberately manufactures for someone else to endure