The Discovery of Anchoring
Tversky & Kahneman (1974) showed participants a number generated by a rigged roulette wheel (either 10 or 65), then asked them to estimate the percentage of African nations in the United Nations. Despite the roulette number being entirely irrelevant to the question, the median estimate from the group that saw 10 was 25%, while the group that saw 65 estimated 45%. An arbitrary number served as a cognitive "anchor" that systematically biased subsequent judgment.
Anchoring is a cognitive bias in which an initially presented value distorts later numerical estimates. Even domain experts are susceptible: real estate appraisers are influenced by listing prices, and judges are swayed by prosecution sentencing recommendations, as demonstrated in controlled experiments. Knowledge and experience reduce the magnitude of anchoring but never eliminate it entirely.
How Rankings Function as Anchors
The moment MyRank displays "You are in the top 30%," that figure becomes an anchor for self-perception. From that point forward, evaluations of one's financial standing or health status unconsciously reference this 30% benchmark as a starting point.
The problem is that the number changes dramatically depending on the comparison group. Being in the top 30% globally might correspond to roughly the median within Japan. The same person in the same condition can appear "above average" or "merely average" depending on which anchor - which reference population - is selected. There is a real risk of locking in a self-assessment based on whichever number happens to appear first.
Anchoring and Goal Setting
Anchoring also distorts goal formation. Someone told they are in the top 30% might set a goal of reaching the top 20%. But this target is itself anchor-dependent. Had the initial display read "top 50%," the goal would likely have been "top 40%." The aspiration shifts in lockstep with the anchor.
A more rational approach to goal setting is to base targets on concrete personal outcomes - improvements in health metrics, increases in savings - rather than on relative ranking positions. "Reaching the top X%" is an anchor-dependent, relative goal that does not necessarily translate into improved quality of life.
De-anchoring - Strategies Against the Bias
Completely eliminating anchoring is impractical, but several techniques can attenuate its influence. First, deliberately consider the "opposite anchor." If shown top 30%, reframe it as "also in the bottom 70%." This alternative framing of the same fact helps restore balance to judgment.
Second, juxtapose multiple reference frames. Comparing one's ranking across different populations - global, national, same age cohort, same occupation - prevents fixation on any single anchor. Third, focus on absolute values. Rather than "top X%," evaluate your situation in terms of concrete figures: annual income in currency units, BMI as a number. Building the habit of thinking in absolutes counteracts the pull of relative anchors.
The Ethical Responsibility of Ranking Tools
Ranking tools influence users' self-perception through anchoring effects. This influence carries ethical responsibility. A display reading "top X%" can induce either pride or inadequacy - both emotional reactions anchored to a number that may not warrant such intensity.
MyRank's practice of presenting contextual information alongside rankings - measurement uncertainty, choice of comparison group, limitations of the metric - is designed to mitigate the harmful side of anchoring. Providing not just a number but also guidance on how to interpret that number contributes to data literacy. A responsible ranking tool does not merely display figures; it equips users with the framework to read them critically.