Cherry Picking

In plain terms

Cherry picking is building a case out of the data points that agree with you and quietly leaving out the ones that don't. The individual facts can all be true. The picture they paint is false, because the facts that would complicate it have been left on the tree.

It's also called the fallacy of incomplete or suppressed evidence. The problem isn't lying. It's curating.

Why it's fallacious

An argument is only as honest as the evidence it's willing to look at. When you select only the supportive data, you guarantee your conclusion before you start, regardless of what the full record shows. The reader is led to believe the evidence points one way, when in fact the evidence was sorted to point that way.

This is what separates cherry picking from a normal argument. Every argument emphasizes supporting points. Cherry picking goes further: it hides the contradicting ones, so the audience can't weigh them.

Canonical example

"This winter was one of the coldest on record in our region. So much for global warming."

One cold winter in one region is a real data point. It's also a single tile pulled from a mosaic of global, multi-decade temperature records that show a clear warming trend. The claim works only if you look at the one tile and ignore the mosaic. Present the full dataset and the conclusion collapses.

The same move powers stock tips ("this fund beat the market last year"), diet claims ("this person lost 40 pounds on it"), and a thousand other arguments. One favorable slice, the rest hidden.

Counter-example (not a fallacy)

"I'm not claiming this is the whole picture. Across the full dataset the effect is small, but in this specific subgroup it's large and consistent, and that subgroup is who the policy targets."

This isn't cherry picking. The speaker acknowledges the full evidence base, then argues for focusing on a subset and explains why the subset is the relevant one. Selecting relevant data with a stated reason is analysis. Selecting flattering data while hiding the rest is the fallacy.

The line: are you showing the reader the whole tree and explaining which cherries matter, or only handing them the ripe ones?

How to fix it

If you've been linked here, the fix is to go looking for the evidence that would cut against you, and deal with it openly. Strong arguments survive contact with the inconvenient data; they don't hide from it. State the full picture, then make your case within it: "Most of the evidence is mixed, but here's why this part still supports my point." That's far more persuasive than a tidy case that falls apart the moment someone produces the data you skipped.

If you're on the receiving end, ask the question that exposes the move: "What does the rest of the evidence show?" or "Is this representative, or the best example you could find?" A sound argument can show its whole hand.