A school wants to attract new families. Its admissions materials lead with the percentage of students who received offers from prestigious universities. The number is impressive. The impression it creates - a school of consistent, high-level achievement - is not wrong. But it is incomplete.
What the materials do not show: the proportion of students who enrolled in community colleges. The proportion who chose vocational training. The proportion who did not pursue further education at all. These paths may represent a majority of the school's graduates. They simply do not appear in the presentation.
A parent who asks for a fuller picture of post-graduation outcomes will find a more complex reality than the headline statistic implies. A parent who does not ask will form a judgment based on the data that was volunteered - which was selected to create a particular impression.
This is cherry picking.
What It Is
Cherry picking is the selective presentation of information that supports a desired conclusion while ignoring or omitting contradictory evidence. It involves highlighting the facts, examples, or data that favour a particular viewpoint and leaving out information that would challenge or complicate that view.
Cherry picking can be conscious - a deliberate rhetorical strategy - or unconscious, driven by motivated reasoning. A person who genuinely believes a conclusion is true may, without any intent to deceive, naturally recall supporting evidence more readily than disconfirming evidence. The selective picture they present may reflect what they genuinely noticed, not what they chose to hide.
This distinction matters for how you evaluate it, but it does not change the epistemic problem: a conclusion drawn from partial evidence is less reliable than one drawn from the full available record.
The Mechanism: Motivated Reasoning Meets the Availability Heuristic
Two tendencies combine to make cherry picking a default rather than an exception.
Motivated reasoning refers to the tendency to evaluate evidence in a direction that supports a preferred conclusion. Evidence that confirms the conclusion feels stronger, more relevant, and easier to remember. Evidence that undermines it gets scrutinised more carefully, filed as an exception, or simply forgotten.
The availability heuristic means that vivid, concrete examples - the ones that easily come to mind - dominate judgment. A single striking success story crowds out the less memorable but statistically more representative majority.
Together, these tendencies mean that the evidence a person volunteers when making a case is likely to be systematically skewed toward what supports their preferred conclusion, even when they believe they are being fair.
Recognising the Shape of Cherry-Picked Arguments
Cherry picking tends to have a recognisable structure:
- A striking positive example or data point leads the argument
- Statistics, if present, are drawn from a narrow window of time or a carefully selected reference period
- Counter-evidence is acknowledged briefly, framed as exceptional, or simply absent
- The conclusion feels self-evident given the evidence presented
The problem is that this structure is also the structure of many genuinely strong arguments. Good evidence is often vivid, specific, and selective by necessity. The question is whether the selection accurately represents the full picture or whether it was chosen to create an impression that the full picture would not support.
Where It Appears
Annual reports and company communications. Companies typically lead with growth and successful initiatives. Missed targets, restructured divisions, and legal challenges appear in smaller type, in later sections, or not at all. An investor relying only on the headline narrative would form a different view than one who read the full filing.
Product reviews and marketing. Hotel websites show attractive rooms. Product pages feature the best-performing specification. Advertisement imagery depicts optimal conditions. The absent information - ordinary rooms, typical conditions, the less flattering specifications - does not appear, because it was not selected.
Anecdotal evidence. A single positive user story, presented as representative of general experience, is cherry picking whether the storyteller intends it or not. The story may be entirely true and entirely unrepresentative of the typical outcome.
Cherry picking leads to overconfidence in judgments and poor decision-making when the omitted information would have been relevant to the conclusion.
See Cherry Picking in everyday decisions
Pick a life context to see how this bias can show up outside the textbook.
Leadership Celebrates Early Praise While Overlooking Usability Issues
A product leader presents a new feature rollout as a triumph based on early user praise, but omits reports of usability problems and support spikes.
Scenario
The product team launches a redesigned checkout flow. Within the first three days, a handful of power users post praise in the internal Slack channel #product-feedback, saying the new flow feels faster. The leader prepares a slide for the executive update, showing screenshots of those Slack messages and quoting the enthusiastic users. When asked about broader metrics, the leader mentions only the positive feedback and does not mention the rise in support tickets logged in the Zendesk queue, the drop in completion rates observed among occasional shoppers in the funnel analytics, or the usability test results that highlighted confusion with the promo code field. The decision to move forward with a full-scale rollout is made on the basis of the selective positive narrative. Later, in a retrospective meeting, the omitted data surfaces, causing a rollback and a discussion about trust between product and support teams.
Where The Bias Enters
The leader focuses on confirming positive feedback because it aligns with the goal of showcasing innovation, while downplaying contradictory evidence due to discomfort with admitting shortcomings.
Decision Check
Before claiming success, review the support ticket trend line and completion-rate funnel for the same period; if either shows a negative shift, request the full dataset including usability test results.
This scenario is illustrative. It explains the pattern and does not claim a documented public case.
Sources
- Rolf Dobelli - The Art of Thinking Clearly