Correlation is not causation - do windmills cause wind

A field of windmills starts turning just as the wind picks up. If all you had was the timing, you could tell a tidy story: the machines are making the air move. The story is clean, visible, and wrong.

The Direction Is the Trap

The mistake is not noticing a pattern. The mistake is stopping there. Wind and windmills rise together because one supplies energy and the other reveals it. The visible motion is an indicator, not the engine.

That is the core mechanism behind Correlation Is Not Causation: two variables can travel together while the real force sits elsewhere. Sometimes A causes B. Sometimes B causes A. Sometimes a hidden condition causes both. Without a causal mechanism, the surface example becomes a shortcut pretending to be an explanation.

Here the moving parts are simple: wind speed changes, turbine blades respond, the observer sees synchronized motion. The observer’s mind then compresses sequence into cause. That compression is useful for speed, but dangerous for truth.

With only the windmill example supplied, keep the model narrow: when a belief rests on “these things happen together,” ask whether you are looking at the source, the symptom, or a shared downstream effect. That is where related ideas like Confounding Variables and Causal Mechanism do real work: they force the hidden middle back into view.

The limit is that correlation is not useless. It is often the first clue. You should not discard patterns just because they are not proof. The failure comes when you treat the clue as the mechanism.

Tomorrow, take one confident explanation you use often and write it as a chain: “X changes Y because…” Then name the physical, social, or incentive pathway connecting them. If the chain has a gap, do not act as if the windmill made the wind.