A Blind Spot at the Heart of Modern ‘Evidence-Based’ Therapy

If you strike ten billiard balls in precisely the same way, you get the exact same result each time.

Human psychology does not work like this.

Say the same words to two different people and you’ll often get completely different responses.

Say the same words to the same person in a different mood and you may get a different result again.

Tone of voice, timing, context, non-verbal cues, mood, expectations, and countless other factors all influence the outcome.

The physical sciences have flourished by treating the world as a machine: find the active ingredient, deliver it efficiently, and the ‘messy’ uniqueness of the person barely enters the equation.¹ 

So it’s understandable that the field commonly called ‘mental health’ has tried to follow this same path.

Pills, step-by-step protocols, and ‘evidence-based treatments’.

It’s a seductive story.

But it’s a deeply limiting story.

Here are three areas of research that begin to explain why.

1. The Doctor Prescribing Can Sometimes Matter More Than the Pill

In the 1980s, the US National Institute of Mental Health ran one of the most famous depression trials ever conducted. Patients received either an antidepressant or a placebo (a sugar pill), handed to them by one of a number of doctors.

For years, the debate centred on the obvious question: do antidepressant medications actually beat placebo? Then, in 2006, Bruce Wampold, one of the most influential outcome researchers in the world, along with Kevin McKay and Zac Imel, went back to the same data and asked a different question.²

Does the specific doctor who prescribes the exact same pill matter?

The answer surprised a lot of people.

On one standard measure the researchers used, the specific doctor handing over the medicine accounted for nearly three times as much of the patient’s improvement as the pill itself.

Even with a pill, the easiest treatment of all to standardise, the human relationship powerfully shapes outcomes.

2. Human Connection Changes Your Biology

You might accept that human connection influences something as subjective as mood.

But what about your immune system?

In 2017, researchers at Stanford pricked volunteers’ skin with histamine, causing an allergic welt.³

Then they applied a placebo cream and told people it would calm the reaction.

When the provider was warm and ‘clearly competent’, the resulting welt ended up being measurably smaller. When the provider was cold and seemed unsure of themselves, that same fake cream did almost nothing.

The quality of the human relationship changed the size of a measurable allergic reaction.

3. The Best Therapists Get Results Ten Times Faster

Decades of research keeps finding that no brand of standardised mainstream therapy reliably beats any other. Psychoanalysis, Cognitive Behavioural Therapies, and dozens of other standardised approaches tend to get the same average results.

This pattern is so robust that researchers have named it the Dodo Bird Verdict, after the Dodo in Alice in Wonderland who declares: 

‘Everybody has won, and all must have prizes.’

If the specific method were the active ingredient, we wouldn’t expect that.

John Okiishi and his colleagues at Brigham Young University analysed 1,841 clients across 91 therapists.⁴

They looked at the school of therapy the therapist practised and how they’d been trained. Neither predicted how well clients actually did.

Instead, what predicted outcomes was something far less comfortable for a manual-based field: the unique therapist.⁵

In fact, clients seeing the most effective therapists improved at roughly ten times the average rate.

And the therapists whose clients improved slowest were also the ones with the most clients who got worse, even though their credentials and methods were no different on paper from anyone else’s.

And this was just from a study of 91 therapists.

Now, imagine what the best in the world can do.

The dream was to scale therapy like a fast-food franchise: identical everywhere, no matter who’s behind the counter. The trouble is that people don’t work that way.

What This Means For You

We are nothing like billiard balls.

We all have our own unique histories, beliefs, fears, and patterns.

This is why the same method, delivered the same way, will never get optimal results for each person⁶. But method is only part of it. The outcome research tells us that who helps you matters enormously.

The most effective practitioners aren’t following a standard method more faithfully. Instead of playing sheet music by rote, they compose, on the spot, the exact music that makes their audience of one come to life.

It’s a refined skill: calibrating continuously to this unique person, with enough precision to create the conditions for fast, automatic change.

And like every refined human skill, it isn’t distributed evenly.

We understand this everywhere else.

There is a vast difference between an average runner and an Olympic champion, between someone who follows a recipe and a Michelin-starred chef, between a teacher who covers the syllabus and the one whose lessons you still remember thirty years later.

However, in ‘mental health’, we’ve been taught to pretend the person helping barely matters, as long as they’re faithfully and competently using an ‘evidence-based treatment’.

The research says otherwise.

None of this is the fault of individual therapists. Most are dedicated people handed a deeply impoverished paradigm (the blind spot in this post being just one part of it).

And the field keeps doubling down: more detailed manuals, more rigid protocols, and now AI, which is wonderful for information, and even for feeling understood, but not for the real-time, deeply human, calibrated experiences that transform automatic patterns.

How to Find the Right Person

So if method, credentials, and experience tell you little, how do you find one of these rare practitioners?

Unfortunately, the research shows it’s not easy to predict from the outside.

The most important thing you can check is this: do they track results?⁷

It’s so easy for even honest, well-meaning therapists, coaches, and other helpers to be overconfident if they don’t.

Why?

One factor is confirmation bias, the natural tendency we all have to notice the evidence that confirms our beliefs rather than the evidence that refutes them.

A large body of research shows that therapists across every school reliably rate themselves as above average, overestimate how many of their clients improve, and underestimate how many get worse.⁸

Another factor is that, in almost any field, knowing what is and isn’t working is essential to getting better at it. Research shows that the average therapist doesn’t improve with experience, and some actually get worse. The exceptions? Those who track results and learn from them.

Unfortunately, even something as basic as tracking results is rare. I’ve had several clients who’d worked with world-famous therapists. It amazed me that not one of those experts had followed up after the sessions to find out whether what they’d done had actually worked (it hadn’t, which was why they’d come to me).

So make sure whoever you work with actively invites honest, accurate, real-world feedback.⁹ Research led by Scott Miller keeps finding the same thing: therapists who track their results and invite honest feedback get better outcomes, and keep improving over time.

Another thing worth looking for is skin in the game. This doesn’t come directly from mental health research, but from everything we know about human behaviour. When the person helping you has little to lose by not getting results, their incentives aren’t aligned with yours. But when someone does have skin in the game, if your time together doesn’t deliver results, the financial cost sits with them, not you.

This is one reason why, for over 25 years, I’ve worked with a unique no change, no pay guarantee. I don’t know anyone else who does this, and, yes, it’s a deliberately high bar. But it’s been crucial to mastering my craft.

You are not a billiard ball.

If change hasn’t happened before, maybe the problem was never you. Perhaps you just haven’t yet found the right person.

Now you know what to look for.

Footnotes for the Scientifically Curious

1. Physics itself long ago moved past simple billiard-ball mechanics. Chaos theory and emergence show that even physical systems can behave in ways simple reductionism cannot predict. My point in this article is narrower: the field of mental health has borrowed the medical ideal of isolating an active ingredient and standardising its delivery. That works beautifully in many domains. It becomes limiting when the individual ‘messiness’ it discards is not noise, but signal. I’m not making the case that minds are too complex to study. I believe precisely the opposite. The claim is only that one- or few-variable RCTs, for all their power elsewhere, are too coarse-grained to catch what matters most here. ↩︎
2. McKay, K. M., Imel, Z. E., & Wampold, B. E. (2006). Psychiatrist effects in the psychopharmacological treatment of depression. Journal of Affective Disorders, 92(2–3), 287–290. ↩︎
3. Howe, L. C., Goyer, J. P., & Crum, A. J. (2017). Harnessing the placebo effect: Exploring the influence of physician characteristics on placebo response. Health Psychology, 36(11), 1074–1082. ↩︎
4. Okiishi, J. C., Lambert, M. J., Nielsen, S. L., & Ogles, B. M. (2003). Waiting for supershrink: An empirical analysis of therapist effects. Clinical Psychology & Psychotherapy, 10(6), 361–373. ↩︎
5. None of this is mystical, as if some people are magically gifted. The difference that matters is simply one that standard research methods are too coarse-grained to detect. A randomised controlled trial finds whether something works on average across many people, which discards the individual detail. Yet that detail is often where the greatest leverage lives. Milton Erickson worked a different way. Instead of relying on a statistical epistemology, he calibrated to the smallest physiological signals, from shifts in breathing to changes in the blood flow in someone’s face. He noticed, for example, how even a slight pause, a certain look, or a single word, delivered with precise vocal quality to the specific individual he was with, could change everything. These idiographic guideposts, loosely informed by nomothetic generalisations, were what showed him what helped the unique person in front of him. I follow in his footsteps. ↩︎
6. Something that in reality can never be accomplished, even when fully scripted. In any two-way human relationship, vocal tone, non-verbal signals, timing, the quality of the interaction, and countless other highly influential factors can’t be controlled. ↩︎
7. Tracking your individual results matters more than any research finding based on groups. Gordon Guyatt, who coined the term ‘evidence-based medicine’, put it this way: ‘Evidence about groups of people can only tell you about groups of people… the best would be to find out what the effect of the treatment is in this patient.’ He places rigorous evidence about the individual at the very top of his hierarchy of evidence, above the multi-person randomised trials normally regarded as the highest form of proof. The best evidence is whether the approach actually works for you. ↩︎
8. In one study, when researchers asked 129 therapists to rate their skill against their peers, the average therapist placed themselves in the top 20%. None rated themselves below average. ↩︎
9. The best mainstream tracking, such as Scott Miller’s, measures broad, self-reported wellbeing, asking you to rate how you’ve been doing across a few general areas of your life, such as your personal life, your close relationships, and your work. That can be useful, for a coarse-grained overview. But in my experience clients tend to care about very specific real-world results and felt shifts no general scale can capture. If panic strikes when you fly, the real question isn’t whether a wellbeing score moved. It’s this: when you next fly, do you naturally feel the way you want to feel? Is flying now as ordinary and easy for you as it is for anyone else?  ↩︎