Cherry-Picking

Cherry-picking is the practice of selecting only the evidence that supports your pre-existing conclusion while ignoring evidence that contradicts it. Unlike confirmation bias — which operates unconsciously — cherry-picking often involves a deliberate or semi-conscious choice to present a misleading subset of the data. A startup reports only its best cohort metrics. A manager highlights the quarters where their strategy worked and omits the ones where it didn't. An investor cites their winning bets and forgets the losers. Cherry-picking is dangerous precisely because the selected evidence is real — it's not fabricated, just incomplete. The cure is disciplined base-rate thinking and the habit of always asking: what am I not seeing? What data was available that didn't make it into this presentation?

Build a discipline of asking for the full dataset, not just the highlights. When someone presents supporting evidence, ask: what's the base rate? How many examples did you look at? What did the failures look like? When building your own arguments, force yourself to include the strongest counter-evidence. If your case can't survive the best opposing data point, it's not a strong case — it's a cherry-picked one.

Cherry-picking selects only the evidence that supports a pre-existing conclusion, creating misleading impressions from real data. In business, it distorts pitch decks, strategic reviews, and performance evaluations. The antidote is disciplined base-rate thinking and the systematic habit of seeking disconfirming evidence.