When Accuracy Gets in the Way of Clarity
“All models are wrong, but some are useful.”
- George Box
In my days as a science teacher, I learnt that pretty much everything we use to explain the world is a model. And as the statistician George Box famously said: “All models are wrong, but some are useful”.
Take the atom. When explaining the model of the atom to a university physicist, I’d talk about probability clouds, where electrons have a certain chance of existing at any one time. This model is helpful when you’re considering how different molecules interact. But for a simple understanding of what makes up our world, it’s totally unnecessary.
With the younger kids, I’d liken the centre of an atom (the nucleus) to the sun. And I’d liken the smaller particles orbiting around it (electrons) to planets. That model is perfect for conveying the idea that the atom is mostly empty space, with its mass concentrated in the centre. It’s not great for understanding the mechanisms of more complex chemical reactions, and that’s fine. When it comes to that, I’d choose a different model.
As an expert, you earn the right to simplify. And when you’re explaining to a non-expert, that’s essential. They don’t have the rich domain knowledge you have. You don’t need to “dumb things down”, but you do need to choose a model that’s fit for purpose. In my time as a physics writer, I’ve written, reviewed and produced hundreds of animated videos. My most common piece of feedback to other writers (and the most common piece of feedback I’ve received!): “Clarity beats accuracy”. This phrase, which I first heard in Dominic Walliman’s brilliant TED talk “Quantum physics for 7 year olds”, is the most effective tool I’ve found to combat expert-induced blindness.
As experts in our field, whatever that may be, we feel responsible and passionate for communicating everything there is to know about it, in the most accurate way possible. That’s great if you’re talking to another expert. But if you’re not, it at best dilutes your message, and at worst loses your audience before you’ve even started. Put another way: “Don’t let the truth get in the way of a good story.”
When we introduce people to unfamiliar phenomena that they can’t picture, we should start in terms they can understand and visualise: show don’t tell. For example, don’t say “Quantum physics deals with discrete energy levels”; say “Think of a staircase – you can stand on one step or the next, but never in between. That’s how energy works in quantum physics.”
Walliman also warns against going “too far down the rabbit hole.” When you’re explaining a complex idea, you don’t just need to know where to start; you also need to know where to stop. That means sacrificing some details so that the most important ones stick.
These principles don’t just apply to physics. Let’s look at some other examples:
How does insurance work?
Insurance works by pooling risk across multiple policyholders, allowing for actuarial calculations that predict probability distributions of future claims.
Accurate. But the version below is clearer.
Insurance is like a safety net that everyone helps build. Everyone pays a little, so if something bad happens to one person, they’re covered.
Another example: what factors determine how much you get paid?
Salaries are influenced by multiple factors including market demand, skill scarcity, industry standardisation, bargaining power, and macroeconomic conditions that shift over time.
Accurate, but unclear. This time we don’t need an analogy, we can just ditch the jargon and explain in simple terms:
People get paid more when their skills are rare and in demand. If lots of companies need what you do, they’ll compete to hire you and pay you more.
Later, when the audience is ready, we can refine it:
Of course, things like experience, negotiation skills, and the economy also play a role.
Perhaps the most famous example of choosing clarity over accuracy is the London Tube map. Deliberately distorted, it shows only the underground lines and stations – nothing else – because that’s all you need to navigate the network. A detailed street map is more geographically accurate, but far less useful if you just want to get from A to B.
The same applies to explanations. Sometimes, simplifying first is the only way to make something useful. Every explanation we give is a model, and no model is perfect. So, pick the one that’s fit for purpose.