From supplements to training plans, athletes are running their own experiments. Here’s how to think critically about what’s working, and what might be an expensive placebo.
There has never been more information about how to train, fuel, and recover, and almost no way to know how much of it applies to you. Your feeds are awash in advice on how to structure your next interval workout, how many carbs to slam while you do it, and whether to chase it all down with bicarb, ketones, or an expensive salad powder sold to you by a guy in a cold plunge.
You read a study and start double threshold. You hear a podcast and start choking down 120 grams of carbs an hour. You watch a pro post their morning routine, and by Tuesday, you own creatine, a heat vest, a $30 ketone shot, and a shoe with as much carbon in it as the Tour de France peloton. Training, supplements, fueling, gear, recovery gadgets—it’s all one big rolling experiment, and you are both the scientist and the lab rat.
On some level, self-experimentation is good. Necessary, even. Because there’s no randomized controlled trial on The Effect of Trader Joe’s Scandinavian Swimmers on the High-Altitude Ultrarunning Performance of Zoë Rom: A Qualitative Analysis, we’re all out here applying exercise physiology to our own n of one. So you experiment.
That’s fine, and you are not a rube for doing it. The rube is the runner who tries the new thing, has one good workout or one good week, declares victory, and never once asks whether that was the intervention or just a Tuesday they happened to feel good (and then immediately shares said intervention in their Instagram stories). The difference between an experiment and a vibe is whether you bothered to check. Checking yourself is a skill, and you can learn it.
N of One
In medicine, a study with a single subject is called an “n-of-1” trial. It’s a real, respected design, and it can tell you whether something works for you specifically, which is the only thing you actually care about. It just requires you to run it like a trial and not like a hopeful person whose Ultrasignup score is rapidly outpacing their credit.
Athletes need to experiment more cautiously and think more critically when translating science or vibes into their own training. Here are the traps that athletes fall into, and how to climb out of each, because the main obstacle to a clean experiment is never the protocol or the powder. It’s the athlete running it. As Richard Feynman said, “The first principle is that you must not fool yourself, and you are the easiest person to fool.”
Regression to the Mean
Let’s say you bonk spectacularly in a 50K. You’re rattled, so you overhaul everything: new training block, new supplement stack, new fueling plan. Your next race goes great, and the changes look like a stroke of genius. You’ve got the belt buckle and a $500 bill from The Feed to prove it.
Except a single bad race is one data point, not your fitness. Maybe it was hot. Maybe you went out too fast, slept badly, or effed around and found out about the aid station watermelon. You were going to run better next time, no matter what you changed, because the disaster was the outlier, not the new normal.
That’s regression to the mean. Your real fitness sits at some average, and any given race lands above or below it depending on the day’s luck: heat, sleep, pacing, melon. A spectacular bonk is that average buried under a once-off pile of bad luck, and bad luck that deep almost never stacks the same way twice. So your next race drifts back toward your true number on its own, and whatever you changed pockets the credit.
Same trap, smaller scale: you start magnesium during your worst week of the year: cramping every night, sleeping like garbage, and within days, both improve. Magnesium, hero. But you reached for it precisely because things had bottomed out, which is exactly when they tend to climb back on their own. Maybe the magnesium worked. Maybe your sleep was going to stabilize anyway, and you just watched yourself regress to the mean.
RELATED: What Happens When Ultra Goes Mainstream
The Shiny New Toy Effect
Call it placebo for athletes, and no, that doesn’t mean fake. Belief is a real, measurable force, and in sport it works by changing how hard you’re willing to go. You lace up the super shoes, or knock back the beet shot with the impressive label and the even more impressive price tag, and suddenly you feel like an athlete on a mission. So you push harder, and you believe a little more. Maybe the bicarb really is helping you buffer the burn. Also, maybe, having taken the bicarb, you’re now convinced you can, so you shove yourself into a level of endurance you’d never otherwise have given yourself permission to reach.
Carbon plates can meaningfully help. Nitrates can do something real. But the swagger of the new thing makes you train harder and lean into the discomfort, and the gear or the gel takes the credit for the gusto. From the inside, you can’t tell how much was the product and how much was the version of you who bought it.
And if you’re the runner thinking, sure, other people fall for placebo, but not me, I am uniquely capable of rational self-analysis, sorry suckers, remember that double-blind trials exist for a reason. Belief is strong enough that even researchers, experts in their own fields, aren’t immune to it. If the folks in lab coats had to build guardrails against fooling themselves, the rest of us can be forgiven for putting a little too much stoke in baking soda and beet juice now and then.
None of which means these things don’t work. Plenty do. Carbon shoes, bicarb, and nitrates all have real science behind them: the question isn’t whether they work, but how much of an edge they hand any particular person, and that depends far more on you (your physiology, how you use it, the demands of the specific effort) than on whether you scored an unusually potent batch of SuperBeets.
The point was never that it isn’t real. It’s that from the inside, without something objective to measure against like a lactate meter, a time-to-exhaustion trial, a study so blinded even the researchers don’t know who got what, it’s genuinely hard to say how much of the edge was the product and how much was you, believing.
Recall and Confirmation Bias
You remember the workout where you felt like a demigod on the new plan, and the morning you woke up convinced the creatine was working. What’s harder to remember are the three flat sessions and the week you felt like a damp sock.
Strava is a highlight reel, the kudos reinforce the story, and “heat training transformed me” turns out to be assembled from the two sessions that happened to feel amazing. Your training log, if you keep one honestly, will frequently disagree with your memory. Memory wants the new thing to have worked. The log does not care how you feel. A journal, a training app like TrainingPeaks or Strava, a coach, or any outside eye willing to tell you the unflattering truth: all of it works as a sanity check on whether you’re seeing your new favorite intervention clearly or through a very flattering filter.
The Novelty Effect
The first two weeks of any new block, or any new supplement, diet, or protocol, feel fantastic, and that’s the trap. You’re fresh, you’re motivated, your fatigue is low, and you’re eating and sleeping and recovering with a sense of purpose you absolutely did not have last month. You’ve found the way. Unfortunately, the magic of week one is mostly that you’re motivated to try the thing, and motivated to find a reason the thing is definitely, certainly working. In the case of training, athletes are going to respond to almost any reasonably intentional stimulus.
Then week three rolls in, fatigue accumulates, the novelty wears thin, and you find out whether cold plunging/paleo dieting/taking in a beluga whale’s worth of carbs an hour was doing anything or whether you were just enjoying being a new and improved version of yourself for a fortnight.
Our brains are wired to over-index on the latest and greatest, great news for ancestors who needed to learn fast which leaves were chill to eat and which would reunite them with theirs, and bad news for anyone trying to make an honest call about whether it was the fancy new workout that moved the needle, or the deeply boring fact that they’d been training consistently for a few more months. (It’s almost always the boring thing.)
RELATED: The Importance of Celebrating Microwins
Other People Are Not You
The pros do double threshold, take bicarb, train their guts to 120 grams an hour, and sleep nine hours a night, so it must all work. But the pro is fast because of genetics, a decade of accumulated base, full-time recovery, a physio on speed dial, a coach, and the ability to nap on the shoe company’s dime. They added the trendy stuff on top of all that.
When you copy the intervention without the context, you’re copying the garnish and skipping the meal, sprinkling steakhouse parsley on a freezer-burned patty and hoping it becomes Ruth’s Chris. The right question is never “does this work for elite Norwegians who have a live-in masseur and ample tax write-offs for foam rollers.” It’s “does this work for me, on my four or five runnable hours a week, with my job, my sleep, and my training age.” The answer, often, will not be to buy the expensive supplement.
Sunk Cost
The heat vest was not cheap (reader: I own two). Neither was the altitude tent, the coaching app, or the ketone esters, which could only cost more if they shipped through the Strait of Hormuz.
It’s not just the money: a twelve-week block is a real chunk of a short season, and admitting it did nothing means admitting you torched it. The more an intervention has cost you, in dollars, in months, in suffering, the louder you’ll defend it, because the brain is built to retrofit cost into meaning.
Surely something that hurt this much must have helped. Nobody believes themselves to be a sucker, so every intervention, run through the psychological immune system, comes back out the far side as IT WORKED! THE EXPENSIVE ALGAE EXTRACT DEFINITELY MADE ME FASTER!
So ask the cold question: if this had been free and taken no time at all, would I still be convinced it worked?
How to Actually Test Something
One honest note first. Some of this you can test cleanly and some you can’t (entire industries exist because of this tension). Training is notoriously hard to test, because fitness is cumulative, different systems and stimuli occur on wildly divergent timelines, and while you can un-take a supplement, you cannot neatly un-train to check whether the effect reverses. Time is a variable.
- Change one thing at a time. If you add heat training, double days, creatine, and a new gel in the same block, you’ll know something changed and have no idea which lever did it. One variable, or you’ve learned nothing.
- Decide what “working” means before you start, and don’t make it a race. A race is one data point on a random day, smeared with weather, nerves, and a thousand things outside the test. Not great evidence, and you’ll be primed to experience it more emotionally.
Use something repeatable: time up a known climb, pace at a fixed heart rate, a standard workout, decoupling on a long run, or for a fueling change, how your gut and energy hold at hour three. Write down the result that would convince you, in advance, before you have a stake in the answer.
- Know your noise, and control your conditions as best you can. Everything wobbles on its own: heat, terrain, sleep, fatigue, how much you ate, whether your dog had opinions at 3 a.m. A test on an 85-degree afternoon is not comparable to one on a cool morning. Same route, similar conditions, similar fatigue state, or you’re measuring the weather and calling it fitness.
- For a supplement, run it, stop it, run it again. For training, give it a real block and re-test. The on-off-on test is the closest thing to a gold standard you can pull off at home, and it works beautifully for a pill or a powder: take it for a few weeks, quit, see if the effect fades, restart, see if it returns. Bonus points if a friend hands you unmarked containers so you genuinely don’t know which weeks are real. You cannot do this with a training method, because you can’t un-train, so instead run it for a genuine six to twelve weeks and re-run the exact same benchmark under the same conditions. The number is the evidence. The feeling is just a story about the number.
- Keep an honest log, not a highlight reel. Record the flat days, the RPE, the conditions, the sleep, the gut, not only the sessions worth posting. Strava is marketing. A log is data, and only if you write down the bad ones too. Bonus: phone a friend. People who are not us might be better positioned to spot our bullsh*t.
- Ask “compared to what.” Better than nothing is a low bar. Is your fancy method actually beating consistent, progressive, deeply boring training with a bit of intensity? Is the supplement beating the version that costs a quarter as much, or beating just sleeping an extra hour?
That unglamorous foundation built nearly every result you admire, it’s essentially free, and it does not come with a vest. Most of the time, the honest finding is that the intervention “worked” because it tricked you into doing the basics consistently for a few months. Which is allowed. Just know that’s what happened, so next time you can keep the consistency and skip the $300.
Everything has a cost. If not in dollars, then in attention, time, and energy that could be channeled into more effective interventions. You could spend $90 on powdered greens. Or you could put that time and money into eating a few more vegetables throughout your week.
Unfortunately for enthusiastic self-experimenters, the boring stuff usually wins. Consistent training, progression, sleep, and enough food. The list of supplements with real, repeatable evidence behind them is short and unglamorous (literally five), and most of the rack is expensive hope. The vests and tents and powders are, at best, a small percentage on top of the foundation, and at worst, a very pricey way to feel like you’re doing something.
RELATED: Embracing the Uncool as Trail Running Evolves
RELATED: What Happens When Ultra Goes Mainstream
RELATED: Know Your Engine: The Ultrarunner’s Guide to Physiology Testing
RELATED: 8 Mistakes New Trail Runners Make (and How to Avoid Them)
