Cherry picking is presenting only the evidence that supports a conclusion while passing over the evidence that tells against it. Every individual fact offered may be true; the argument misleads through what it leaves out rather than through what it asserts.
Alternative Names: Suppressed Evidence, Fallacy of Incomplete Evidence, Card Stacking, Fallacy of Exclusion
Cherry picking, also called suppressed evidence, is a fallacy of selection. The arguer surveys a body of evidence, keeps the part that fits the desired conclusion, and presents it as though it were the whole. Because nothing stated is false, the argument can survive fact-checking line by line and still give a thoroughly misleading picture.
This makes it unusually hard to spot from the inside of an argument. The question to ask is not "is this true?" but "what would I need to have seen alongside this?" A single quarter of falling crime, a single study, a single year of temperature data — each can be entirely accurate and entirely unrepresentative of the series it was drawn from.
Cherry picking shades into neighbouring fallacies that are worth keeping distinct. Anecdotal reasoning leans on one story in place of systematic evidence; cherry picking actively suppresses the systematic evidence that disagrees. Texas Sharpshooter chooses the frame after seeing the data, whereas cherry picking chooses the data. Quoting Out of Context misrepresents a single source rather than selecting among many.
Not all selection is fallacious, and the distinction matters. Summarising evidence necessarily means leaving most of it out, and excluding a study on stated methodological grounds is ordinary practice. What makes it cherry picking is selecting on the basis of the conclusion — keeping what helps and discarding what does not, while presenting the result as a fair survey.
