Three weeks after Jess starts taking the turmeric capsules her sister-in-law swears by, the ache in her knee finally settles. She hasn’t missed a dose since, and she’s told half your 6am class about them. Nobody mentions that she cut her running from four days a week to two in that same month.

Jess has fallen for the illusion of causality, and so has almost everyone who has ever bought a supplement. The illusion of causality is the tendency to believe that one thing caused another because the two happened together, even when the outcome would have happened anyway. Your brain counts the times the capsule and the relief lined up, and it never goes looking for what happens without the capsule.

The Psychology Behind the Illusion of Causality

  • What it is: Researchers also call it the causal illusion, and they define it as seeing a cause-and-effect link between two events that are statistically unrelated. When the supposed cause is your own action, psychologists call it the illusion of control, which is the same error pointed at yourself.
  • Who found it: The roots go back to B. F. Skinner’s 1948 “superstition” experiment, in which pigeons fed at fixed intervals, no matter what they did, developed odd rituals like turning in circles, as if their behaviour had produced the food. Ellen Langer showed in 1975 that people who picked their own lottery ticket wanted far more money to sell it than people who were handed one, and in 1979 Lauren Alloy and Lyn Abramson found students badly overestimated their control over a light that switched on regardless of what they did. The modern research programme belongs to Helena Matute and her colleagues at the University of Deusto in Bilbao, who have spent more than 20 years mapping when the illusion appears, with researchers in Sydney more recently extending it to the messier, variable outcomes of real life.
  • The psychology: Judging whether something works takes two numbers: how often the result happens when you do the thing, and how often it happens when you don’t. Your brain is lazy about the second number. It gives extra weight to the moments when cause and result line up, and decades of lab work point to two conditions that inflate those moments. The first is outcome density: when the result is common anyway (colds clear up, sore backs settle, beginners get stronger), any cause looks effective. Cause density is the second condition, because the more often you do the thing, the more coincidences you collect and the fewer chances you get to see what happens without it. Blanco, Barberia and Matute (2014) found this runs as a loop, since a remedy with no side effects gets used more often, and using it more often makes people believe in it more.
  • The key study: Blanco, Matute and Vadillo (2013) gave participants a fictitious drug and a run of fictitious patients whose recovery had nothing to do with the drug. They varied two things independently: how often the drug was given (80% or 20% of patients) and how often patients recovered (80% or 20%). Judgements of the useless drug were inflated whenever recovery was common, and they peaked when both numbers were high, which is the everyday reality of a popular remedy for a condition that fixes itself.
  • The damage it does: Yarritu, Matute and Luque (2015) let one group build a strong illusion that drug A worked, and another group only a weak one. Both groups then saw drug B, which did work, given alongside drug A. The group that believed in drug A rated drug B as less effective, so a false belief about one treatment took credit away from the treatment that earned it.
  • Debiasing: Matute (1996) showed people judge far more accurately when their goal is to test the drug rather than to heal patients, and when they give it to roughly half of them so they can see what happens without it. Barberia and colleagues (2013) cut the illusion among teenagers with one workshop on experimental control, although a 2024 study by Chow and colleagues found scientific-method training gives only partial protection. The fix works in principle and leaks in practice.

How the Illusion of Causality Shows Up in Your Gym

  • In Marketing & Sales: You launch a new ad creative on 2 January, your enquiries triple by the middle of the month, and the creative becomes the hero of your next team meeting. Enquiries triple every January. Your ad ran at the same moment the outcome was going to happen anyway, which is the textbook setup for a causal illusion, and the only honest test is to compare this January with last January or to split your spend so some audiences never see the new ad. Testimonials make it worse, because they only ever show people who bought the product and got results, never the people who got the same results without it. Most gym marketing reports skip that comparison, and plenty of mediocre campaigns get renewed because of it.
  • In Client Behaviour & Retention: Members start a fat burner, a collagen powder or a set of recovery boots in the same fortnight they finally start training three times a week. The product has no side effects, so they use it every day, and daily use is the condition that inflates the illusion most. When the results arrive, the product takes the credit, because it’s the new, visible thing and the consistency isn’t. Yarritu’s work predicts the expensive part for your gym member retention: a member convinced the tub works gives less credit to the sessions that did the real work, and a member who doesn’t credit the training has no reason to protect it when life gets busy.
  • In Personal Training / Coaching: Coaches fall for this harder than clients, because we pick the method and then watch for it to work. Beginners get stronger on almost any sensible programme for their first few months, so every coach’s pet method looks brilliant when the outcome is that common. Pain behaves the same way. A client arrives with a grumbling lower back, you swap her deadlifts for trap-bar pulls, and three weeks later she’s pain-free, but a lot of everyday back pain eases within weeks on its own, which is why Matute’s team single out back pain as one of the conditions alternative practitioners treat most. That doesn’t mean your programme did nothing. It means you can’t tell from one client, and neither can she.
  • In Business Operations / Staff: You trial a new cancellation-save script at one club in November, cancellations dip over the next six weeks, and head office rolls it out everywhere. Nobody checked whether the other clubs, still on the old script, dipped just as much over the same weeks. Seasonal patterns, a price rise at the competitor down the road, or a strong run of new classes can all move cancellations, and any of them can hand your script credit it didn’t earn. The same trap runs in reverse when a new sales hire starts in a slow month and gets written off. Before you back an idea or a person on the strength of one good month, find a comparison that didn’t get the change.

Real-World Examples from Other Industries

  • Power Balance: Power Balance sold silicone wristbands with an embedded hologram that it claimed improved balance, strength and flexibility, and the sales pitch was a live demo: try the balance test without the band, then again with it. The second attempt nearly always looked better because the person had warmed up and knew what was coming. John Porcari’s team at the University of Wisconsin-La Crosse randomised the order with 42 athletes and found no difference between the real band and a placebo. In December 2010, the company admitted to the ACCC that its claims had no credible scientific basis and offered refunds to Australian customers. The demo worked because it manufactured the coincidence the illusion feeds on, right in front of the buyer.
  • Airborne: Airborne, the effervescent vitamin tablet marketed as “created by a school teacher”, told buyers to take it at the first sign of a cold or before walking into a germy place like a plane. In 2008 its former owners paid US$23.3 million to settle a class action over claims it could prevent or cure colds. Those instructions were a near-perfect recipe for the illusion. Colds clear up by themselves, so almost everyone who took Airborne at the first sniffle got better soon afterwards and gave the tablet the credit.
  • Lumosity: Lumosity sold “brain training” subscriptions on the promise that its games would lift performance at work and school and hold off age-related decline. Users watched their game scores climb, felt sharper, and linked the two, even though getting better at a memory game mostly makes you better at that memory game. In 2016 the US Federal Trade Commission charged the company with making claims it couldn’t back with science, and Lumos Labs settled for US$2 million.

Related Biases & Mental Shortcuts

  • Confirmation Bias: The illusion of causality runs on the same habit of testing a belief by looking at the cases where you did the thing and ignoring the cases where you didn’t. It’s why a member who takes a pre-workout before every session never finds out what an unassisted session feels like.
  • Anecdotal Fallacy: One member’s transformation story is a single coincidence dressed up as proof. The anecdotal fallacy is what turns Jess’s knee into a recommendation for the whole 6am class.
  • Outcome Bias: Once you’ve decided a change caused a good result, outcome bias makes you rate the decision itself as smart. Together they make it very hard to drop a tactic that got lucky once.
  • The Self-Serving Bias: When a client improves, coaches credit the programme, and when a client stalls, they blame the client’s sleep or diet. The self-serving bias picks which causal story you tell, and the illusion of causality makes that story feel like evidence.

The “BS” Takeaway

The fitness industry runs on the illusion of causality. Supplements, gadgets, recovery fads, “fat-burning zones” and half the programmes sold on Instagram survive because something good happened around the time people used them, and nobody checked what happened to the people who didn’t.

You can’t switch the illusion off in your members, and you can’t switch it off in yourself. What you can do is refuse to profit from it. Don’t run with-and-without demos that stack the deck, don’t let a supplement rep turn your members’ coincidences into testimonials, and when a member credits a product for results their training earned, hand the credit back to the training. That last habit keeps members too, because people stay loyal to whatever they believe is working.

Twenty years of Matute’s research keeps landing on the same defence: look at the people who didn’t do the thing. Run your marketing, coaching and staffing decisions through that one check and you’ll kill a lot of expensive nonsense.

Most fitness BS dies the moment someone asks what happens without it.

Bibliography & Further Reading