Placebo in training: how much strength comes from expectation

You start a new supplement, two sessions later your bench moves 5 kg and it feels like you have found a shortcut. How much of that is the product, and how much is the fact that you expected it to work?

This is not a philosophical question. Expectation has a measurable effect on how much you lift and how long you last, and it is big enough to ruin any test you run on yourself in a single session. Here is what is solid, what is only plausible, and where the claims get stretched.

What is reliably known

Expectation changes output, not just feeling. When people are told they have been given caffeine, beetroot juice or some new legal performance aid and are actually handed an inert capsule, they average slightly more reps, slightly longer time to exhaustion or slightly higher power. The evidence here is consistent — the finding repeats across different tasks, different labs, and both trained and untrained people.

The effect is real but modest. The usual scale is a few per cent. In endurance work that means roughly 1–3 per cent on time or average power; in strength work it is typically one or two extra reps in a set, or a couple of per cent on a lift. That matters enormously on a platform where places are decided by a single kilo, and it explains almost nothing about why somebody gains 8 kg of muscle in a year.

It works through effort and discomfort tolerance. Perceived exertion in the same set drops a little, the point at which you decide you are finished shifts a little further out, and one more rep happens. You do not get new muscle or a better heart. You get access to a slice of capacity you were routinely leaving on the table.

It is largest where you decide when to stop. Sets to failure, tests to exhaustion, sprints — all of them end because you choose to end them. Where output does not depend on your decision, such as a muscle contracted by electrical stimulation, the expectation effect largely disappears. That is one of the cleanest findings in this area and the best marker of where placebo stops.

Nocebo is real and no weaker. Tell someone they have received the inferior version of a drink, or that they slept badly, or that their form looks off, and performance drops — sometimes by more than a positive message lifts it. The evidence is solid here too, though the negative side has been studied less than the positive one.

What is likely but not settled

Almost everything we know confidently comes from a single session or a single test. What expectation does across months — whether someone who believes in their programme genuinely builds more muscle and strength than someone who does the same programme sceptically — is poorly researched. There simply are not enough long trials with that design.

The reasoning is sensible: more effort per set, fewer skipped sessions, more work accumulated over 12 weeks. But that is an inference from short-term results, not a measured outcome. Treat it as such.

Nor is it known how much of the spread between people on the same plan belongs to belief. Sleep, nutrition, training history and biology account for a great deal, and how much is left for expectation is unresolved — one reason the question of why two people on the same programme progress differently is still open.

Open-label placebo, where you are told upfront that the capsule contains nothing active and it still does something, has been documented mainly for pain and fatigue in clinical settings. In sport there are only a handful of small trials with mixed results. Interesting, too thin for advice.

The same applies to spotting who responds. No personality trait, no measure of suggestibility and no amount of training experience has turned out to predict it reliably.

What gets reported wrongly

“Placebo means it wasn’t real.” The opposite. Five kilos on the bar is five kilos regardless of what moved it. Placebo does not mean imaginary; it means the cause was expectation rather than a molecule.

“Thirty to forty per cent of everything is placebo.” That figure comes from clinical trials on pain and similar subjective outcomes, where it makes sense. It does not transfer to kilos and seconds — for physical output the numbers are much smaller.

“Supplements only work because of placebo.” Close to true for most of the shelf, but not for all of it. A few hold up when nobody knows who received what; the clearest example is what creatine actually does, and caffeine has a measurable physiological effect alongside expectation, which is also why caffeine tolerance behaves the way it does.

“Once you know it’s a placebo, it stops.” Stated far too confidently. The evidence is inconsistent and mostly from medicine rather than the gym.

Every good day gets credited to placebo. Most good sessions are ordinary variation: better sleep, a warmer gym, less stress, a different exercise order. That is exactly why one session proves nothing in either direction.

What this means for your training and nutrition

Keep the ritual, don’t pay for it. Same warm-up, same order, same playlist, the same ten minutes before your first working set. It is the cheapest way to use expectation and it requires no product.

Guard against nocebo harder than you chase placebo. “I slept badly, today is a write-off” said in the changing room tends to come true. A rule that works: do the warm-up and the first working set, then decide whether to cut the session short.

Blind yourself where you can. Have a partner load the bar without telling you, stop reading the plates, run some sets by effort rather than by number. Reps in reserve is a useful tool for that, and the article on how close to failure to train covers how to use it.

Never test a supplement in one session. Give it 4–6 weeks minimum, same exercise, same time of day, every set written down. Anything shorter measures your mood, not the product.

Expectation shapes soreness too. If someone tells you a session will wreck you, you are more likely to report worse soreness afterwards. Part of that is genuine physiology, explained by the repeated bout effect, and part is pure anticipation.

Spend by evidence, not by packaging. Price, flavour and label design all raise the placebo and none of them raise your twelve-month result. What placebo cannot substitute for is progressive load, enough protein and turning up consistently.

When to see a doctor

Placebo acts most strongly on the experience of pain, which makes it dangerous when pain is the signal you need. See a doctor if pain lasts beyond two or three weeks without responding to rest, if it wakes you at night or is present at rest, if there is swelling, joint instability, numbness or weakness, or if you get chest pain, dizziness or fainting during exertion. No belief and no ritual is a reason to push through any of that.

In short

• Expectation genuinely changes output, but modestly: usually a few per cent, or one or two extra reps.
• It works by lowering perceived effort and moving the point where you decide to stop; it vanishes where the decision is not yours.
• Negative expectation costs at least as much as positive expectation gives.
• The long-term effect on muscle and strength has not been studied well enough to claim anything about it.
• Rituals are free, and one session is never a test — of a supplement or of a programme.