Why science gets it wrong (and why that doesn’t mean it’s broken)

Science can be frustrating—progress is often slow, and research we trusted can turn out to be incomplete or even wrong. That makes it hard to make informed decisions about things like your health or treatment options.

If you’ve ever asked “How could the experts get it wrong?”, maybe this post will help explain how scientists are doing the best they can, and why the process—while flawed—isn’t failing.

Starting with nothing

Scientific experiments are usually designed around a null hypothesis. Yes, that already sounds boring—but it’s important. The null hypothesis is just a statement that there’s no effect or no relationship—that nothing interesting is going on.

Why start with the assumption that nothing exciting is happening? It’s not to suck the joy out of science—it’s to reduce accidental bias. Scientists love their work, and it’s easy to start seeing patterns or connections that aren’t really there. The null hypothesis is like a guardrail: it keeps conclusions grounded in evidence rather than enthusiasm.

Instead of looking for proof that something is happening, scientists gather evidence to reject the idea that nothing is. And even if they can’t gather enough evidence, that doesn’t mean the null hypothesis is true—it might just mean they didn’t measure the right thing, or didn’t measure it accurately enough.

Science is more like a courtroom drama—a “beyond reasonable doubt” situation. Why? Because it often takes a landslide of studies and repeated evidence to truly prove or disprove anything unequivocally. And even then, our understanding of the world is constantly evolving.

Why experiments can seem so weird

At first glance, some experiments seem downright weird or disconnected from reality. But it’s often because the real meaning—why the detail matters—is buried in previous research, deep in a dense academic paper, or tangled in layers of concepts that separate it from its practical use.

Let’s focus on biology, pharmacology, and physiology—fields where a lot of labwork or real-life studies, deal with complex, living systems—people, animals, and even plants.

Layers of complexity

Everything is made up of nested layers of complexity, like a Russian doll:

Bodies → Organs → Tissues → Cells → Organelles → Molecules → Atoms → Subatomic Particles → Empty space.

Yes—even nothing matters.

In this kind of system, any change—even stopping or removing something rather than adding it—can ripple outward and affect countless other processes. And whatever you’re observing can also be affected by multiple other ripples, started by countless other changes, all going on at the same time.

Controlling the variables

That’s why experiments try to control as many other influences as possible. Scientists talk about “controlling the variables”—meaning they try to hold everything else steady so they can be sure that X caused Y.

But the further up the complexity scale you go—from molecules to whole people—the harder that gets. In a test tube, you can control temperature, light and humidity. In a human body? Not so much. Food, exercise, illness, environment, ethnicity, age—even your DNA is a variable. In a single study, there could be multiple uncontrollable variables affecting one person, or different variables affecting multiple people. No wonder mad scientists are portrayed with wild hair—they’re probably just pulling it out.

What’s even more frustrating is that, in real life, some changes only matter because they usually happen alongside something else. But in the name of scientific clarity, researchers often separate and isolate those elements. Ironically, that effort to simplify can strip away exactly what makes the change meaningful in the first place. The desire to be sure sometimes means we can’t be sure at all.

“The desire to be sure sometimes means we can’t be sure at all.”

– on the paradox of scientific precision

Right and wrong at the same time

That’s why science can sometimes be right and wrong at the same time. A study might be well-designed, properly run, with solid data and analysis—but still miss something crucial.

Why? Because no one knows to look for it yet. There might be an “X factor”—a molecule, a process, a feedback loop—that affects the outcome, but hasn’t been discovered.

Science is limited by what it can measure. If we can’t measure it, we probably don’t know it exists. And if we don’t know it exists, we can’t build tools to measure it. That’s the paradox: science is best at investigating what we already suspect is there.

Accidental discoveries and educated guesses

Many breakthroughs happen by accident. Sure, scientists pursue specific goals—like new antibiotics or ways to detect disease—but they’re often building on what’s already known, using predictions and educated guesses. This is the realm of known unknowns.

But what about the unknown unknowns—the mysteries we don’t even realise are there? I can’t begin to imagine how much of life’s complexity still falls into that category. It’s exciting. It’s also a little terrifying.

Faith in science—or in ourselves?

Science takes a kind of faith—not in the results themselves, but in the process. A belief that answers are out there, and that we should keep looking.

The problem is, we sometimes put our faith in our interpretation of science instead of the science itself—and that’s when things can go wrong.