Proof Under Exertion: What Pull-Ups to Failure Revealed About Human Output
Everything in this series has argued the same physiological mechanism: heat accumulates in the body, core temperature rises, and performance falls away to protect it. Cool the core and you give work capacity back.
This piece is where that argument meets a bar and a stopwatch. Pull-ups to failure are a brutally honest test — you either complete the rep or you don't, and there is no way to fake the count. Which is exactly how what is arguably the most important observation in this technology's history happened, entirely by accident.
The Accident: A Protocol That Wasn't Supposed to Measure Performance
The Stanford team wasn't studying strength. They were studying heat extraction, and to measure it they needed an overheated person in a controlled environment. A lab assistant who exercised regularly and intensely volunteered to do his workouts in the lab, and they created a fixed protocol around him: ten sets of pull-ups to muscle failure, with three minutes of rest between sets. The point was to drive his core temperature up so they could measure how much heat their device could pull back out. Performance was not the experiment.
Then, one day, something happened that wasn't in the protocol.
Sit with the logic there. After ten sets to failure, this subject should have had nothing left. Instead, once cooled, he matched his opening set. The fatigue that had been ending his sets was not primarily damage or depletion — it was substantially thermal. Remove the heat and the performance ceiling moved.
That single unplanned observation reframed the entire technology: from a way to cool people down to a way to restore what heat had taken.
The Magnitude: 180 to 618
Having seen it once, the team did the obvious thing — they started cooling him between sets rather than only at the end, and tracked what happened to his total work volume over time. The progression is the most striking result in the CoolMitt library, and it comes in two halves.
training normally
cooling after every other set
First six weeks, training normally. He began at roughly 100 pull-ups per session and worked up to about 180 by the end of the period. That is a real, respectable training adaptation — the kind of progress hard, consistent work produces.
Second six weeks, cooling after every other set. He went from 180 to 618 pull-ups in a single session.
That's not a marginal gain. It's a different training life. And the interpretation Heller draws from it is the part that matters for anyone who works or trains under heat:
Cooling didn't make the muscle stronger on its own. It removed the thermal ceiling that was capping how much work the muscle could do — and the additional work is what drove the conditioning. That distinction is the entire mechanism, and it's why this transfers directly to a worker's shift: more quality work per session, at the same effort.
The Replication: What Happens Set by Set
The lab result raised an obvious question — is the benefit spread evenly across a session, or does it concentrate somewhere? Ricky Doyon, a strength-and-conditioning expert, ran a structured self-test on exactly that, doing sets to failure in consecutive weeks with and without palmar cooling between them.
His counts, week over week:
| Set | Without cooling | With cooling | Change |
|---|---|---|---|
| 1 | 20 | 22 | +10% |
| 2 | 14 | 19 | +35.7% |
| 3 | 11 | 16 | +45.5% |
| 4 | 8 | 18 | +125% |
| Total | 53 | 75 | +41.5% |
Look at the shape of that data rather than the headline number. On the first set — when he was still fresh and had accumulated almost no heat — cooling had little impact: two extra reps. By the fourth set, when heat had fully accumulated, he more than doubled his output.
That's precisely what the mechanism predicts. If the fatigue being removed is thermal, the benefit should be small when there's little heat to remove and large when there's a lot. The effect grows exactly as the thermal load grows. A result that scaled evenly across sets would actually be harder to explain.
In the Field: Breaking a Three-Year Plateau
Away from the lab, the pattern shows up in ordinary training logs. One powerlifter who tracks every session described being stuck for years before adding cooling between sets:
On the squat, over roughly two months of use, his best front squat went from 15 reps at 465 lb to the same 15 reps at 495 lb — thirty pounds added to a lift he considered maxed out. His own description of it: "seeing jumps like that is just unheard of."
His account of how he uses it is unglamorous and worth repeating, because it matches the Stanford protocol exactly: about a minute and a half of cooling between sets, extending to roughly two minutes before his heaviest, all-out set — the one where the accumulated heat would otherwise cost him the most.
What This Proves — and What It Doesn't
This is the point in a marketing article where the temptation is to call all of that "clinical proof." It isn't, and saying so would undercut everything else this series has argued.
Be precise about what these results are. The Stanford pull-up progression is a single subject observed over twelve weeks. Doyon's test is a structured single-subject comparison across two weeks. The lifting results are self-reported training logs from an individual user. These are anecdotal observations — striking ones, consistent with each other, and consistent with the underlying physiology — but they are not randomized controlled trials, and they carry all the usual caveats: no control group, no blinding, and ordinary training progression that can't be fully separated out.
(That being said, the data is consistent with peer-reviewed, published studies and are all from credible athletes.)
Now be precise about what the numbers do show. Across independent settings — a Stanford laboratory, a strength coach's structured self-test, and a powerlifter's training log — the same signature keeps appearing: when heat is removed between bouts, work volume goes up, and the effect grows as thermal load accumulates. That's the exact pattern the physiological mechanism predicts, showing up in people who weren't looking for it, measured by counts that are hard to fudge. The subject after ten sets to failure did not feel less tired and estimate more reps. He did them, and someone else recorded the results.
That convergence is what makes the mechanism credible. The individual numbers are evidence, not proof — and evidence pointing consistently in one direction, from sources that don't share a method, is how confidence in the veracity of the science is properly built.
Why the Bar Matters to People Who Never Touch One
It would be easy to file this under athletics. That would miss the point.
A pull-up to failure is just a clean way to measure something every working body does: produce output until heat forces it to stop. The firefighter climbing stairs in turnout gear, the lineman on a hot pole, the warehouse crew in the seventh hour of a double — all of them are running the same protocol as that lab assistant, in worse conditions and without a three-minute rest.
What the bar proved is that the ceiling those workers hit is partly a thermal one, and thermal ceilings can be moved. That's why this piece sits in the middle of a series aimed at industry: the athletes hit the wall first and most measurably, which makes them the clearest place to observe a limit that everyone shares.
Sources: Stanford University — palmar cooling research by Dr. Craig Heller and Dr. Dennis Grahn ("The Eureka Moment: Cooling Increases Performance"), CoolMitt science library · Structured pull-up comparison by Ricky Doyon, strength-and-conditioning expert · Squat and bench results from a CoolMitt user's self-reported training logs. Results described are individual case observations, not controlled clinical trials, and are presented as such.