The Barkley Marathons has always defied logic. It’s an ultra where failure is the norm, and a legendary “victory”—if it happens—is, well, just finishing.
Held each spring in the rugged terrain of Frozen Head State Park, Tenn., the Barkley challenges 40 runners to complete five loops of 20-ish miles (for a total of roughly 100 miles) through unmarked, mountainous trails. It includes self-navigation with map and compass, a scavenger hunt for book pages to prove loop completion, and an evolving set of maniacal obstacles that shift each year. Each loop has strict cutoffs, and missing any means disqualification. In nearly four decades, only 20 people have ever finished the race.
In 2025, the race once again reminded the ultra world of its mythic brutality.
This year’s edition saw zero finishers, just one “Fun Run” completion (three loops), and only 10 runners out of 40 starters successfully completing even the first loop. It was a collapse in performance so stark that it forced a question rarely asked so explicitly: Was this the toughest Barkley Marathons ever?
The Barkley has always existed as a kind of cold war between the course and the competitors. Each year, runners come prepared to test themselves against a race that seems determined to beat them. And behind that race is its creator, Lazarus Lake, the race’s puppet master. When five runners crossed the finish line in 2024—a record—it was widely understood to be an outlier. The 2024 edition was, according to our performance metrics, the easiest Barkley ever recorded.
Lazarus Lake, never one to tolerate such a swing in momentum toward the runners, likely recalibrated the course for 2025. And the numbers show it. This year was not just a reversion to the mean—it was a dramatic correction. The Barkley struck back.
We crunched the numbers across 25 years of Barkley data and constructed a year-over-year performance analysis. The results are compelling—and paint 2025 as one of the most devastating editions in Barkley’s modern history.
The Fewest Loop 1 Finishers in Modern Barkley History
Historically, Loop 1 is manageable for many of the ultrarunners selected to race the Barkley. In most years, a majority of starters make it through the first loop before attrition begins in earnest.
Not this time.
In 2025, only 10 runners completed Loop 1, marking the lowest finish rate since 2000, the beginning of Barkley’s modern data era.
Graph Explainer: This chart shows the number of runners who completed Loop 1 each year from 2000 to 2025. The sharp drop in 2025 highlights how early and decisively the course dismantled this year’s field. It’s the only year in which Loop 1 finishes dipped to 10 or fewer.
The Second Slowest First Loop Split Since 2000
Even the fastest runner in 2025 struggled. Maxime Gauduin, one of two French competitors to finish Loop 1, posted the fastest time: 9 hours and 45 minutes.
That’s the second slowest leading Loop 1 time on record, only ahead of 2021, when the fastest runner clocked in at 10 hours and 22 minutes.
Graph Explainer: This line chart tracks the fastest Loop 1 finisher in each year since 2000. The higher the number, the slower the leading runner was. 2025’s placement near the top of this graph shows just how much slower front-runners were compared to historic norms.
Scoring the Suffering: The Barkley Performance Score
To quantify yearly performance, we developed two scoring systems.
1. Linear Score (Max 200 Points)
Each loop completed earns points:
- Loop 1 = 1 pt, Loop 2 = 2 pts, and so on, with Loop 5 (or a finish) = 5 pts
- Max score = 200 pts (if all 40 runners finish. Ha!)
2025’s score: 25 pts → 12.5% of max → Lowest since 2000 (the furthest back the Barkley stats go).
Graph Explainer: This bar chart shows each year’s total performance score as a percentage of the 200-point maximum. 2025 is the worst-performing year, even compared to other “zero finisher” years.
Before introducing this second system, it’s worth noting why a simple loop-by-loop score doesn’t fully capture the essence of the Barkley. While finishing any loop is an achievement, completing the latter loops—especially the fifth and final one—is a feat of extreme rarity. To account for this, we created an exponential model that assigns far more weight to the deepest finishes, reflecting the true difficulty of progressing in this race.
2. Exponential Score (Max 640 Points)
Because finishing later loops is exponentially harder, we applied a curve:
- Loop 1 = 1 pt
- Loop 2 = 2 pts
- Loop 3 = 4 pts
- Loop 4 = 8 pts
- Loop 5 (Finish) = 16 pts
- Max score = 640 pts (all 40 finishers; again, ha!)
2025’s score: 66 pts → 10.3% of max → the lowest ever score.
Graph Explainer: This chart reveals the value of deep-loop finishes, especially full completions. 2025’s complete absence of finishes leads to its rock-bottom score, reinforcing its place as one of the most difficult years on record.
The Barkley Brutality Score: A More “Barkleyesque” Way of Measuring the Race’s Merciless Edge
In true Barkley fashion, we decided to invert our performance analysis. If the race is meant to be unforgiving, why not score it based on how thoroughly it crushes competitors? Enter the Barkley Brutality Score — a metric where 100% represents total carnage, and 0% would mean an unthinkable scenario: everyone finishing.
This score flips the Exponential Performance Score on its head, measuring how hard the race was on its entrants. The result? A clearer sense of which years were the most merciless.
And no surprise: 2025 reigns supreme. With a brutality score of 89.7%, it far outpaces most years, and the swing in brutality from 2024 to 2025 is the steepest ever recorded — a 33-point spike in suffering. It’s a fitting reversal for a race known for its pitiless nature.
The swing in brutality from 2024 to 2025 is a 33-point spike — the most extreme shift in race difficulty on record.
So it’s fair to say that, between the weather and Lazarus Lake’s course modifications, as well as the runners’ individual strategic decisions and abilities, this year was clearly a massive correction back to the mean. And then some.
A Case Study in 2025’s Brutality: John Kelly’s Slowest Attempt
Few runners exemplify Barkley’s challenge like John Kelly. He’s made eight attempts, finishing three—most recently in 2024.
Here’s how his 2025 effort unfolded:
- Loop 1: 10:57 (nearly 2.5 hrs slower than 2024)
- Loop 2: 25:00
- Loop 3: 39:50 – he completed the Fun Run (three loops) but missed the 36-hour cutoff required to continue.
John Kelly’s Loop Completion Times by Year
Graph Explainer: This chart tracks John Kelly’s loop completion times across all his attempts. The steep climb in 2025 shows his slowest progression ever, reinforcing how difficult this year was for even the most experienced Barkley athletes.
Experience-Weighted Scoring: How Kelly’s Efforts Stack Up
To further understand Kelly’s trajectory, we applied an experience-weighted exponential score that favors early attempts more heavily. The idea: a runner’s first years are inherently more challenging, so we graded his loop completions against a curve that accounts for that. Finishing deeper loops in early years scores closer to 100%, while later performances carry a lower weight.
Graph Explainer: This chart shows Kelly’s Barkley performances scored out of 100%, with earlier years rewarded more steeply to reflect the difficulty of performing well with less experience. It illustrates both his growth and the challenge of sustaining top results over time.
It Was A Year of Reckoning
The data doesn’t lie. 2025 wasn’t just another zero-finisher year—it was an outlier in terms of difficulty, attrition, and raw performance metrics.
Whether due to course changes, weather conditions, or a combination of both, 2025 reestablished the Barkley Marathons as the world’s most unforgiving ultra. No finishers. No Fun Runs. Just pain—and history.
