Leadership Replacement and Regression to the Mean

Definition

Leadership replacement and regression to the mean is the attribution error in which an organization dismisses a coach or executive after an unusually poor result and treats a subsequent return toward typical performance as proof that the replacement caused the improvement.

Current Synthesis

What Makes Great Leaders & How to Live an Extraordinary Life | Dr. Jared Diamond uses sports and business to separate temporal sequence from causal evidence. Replacement often follows an extreme low, exactly when random variation, injuries, schedule, market conditions, or temporary shocks make some rebound likely even without intervention.

The concept does not imply that leaders never matter or should never be replaced. It raises the comparison standard: evaluate performance against an appropriate no-change baseline, multiple periods, peer organizations, roster or workforce changes, institutional conditions, and the mechanism by which the new leader is supposed to improve results.

Key Claims

• Extreme poor performance often triggers replacement at a statistically favorable moment for apparent recovery.
• Improvement after replacement does not by itself identify replacement as the cause.
• Leader evaluation needs a credible baseline that includes prior variability and external conditions.
• Player, workforce, company, industry, and institutional changes can coincide with a leadership transition.
• Mechanism evidence and repeated performance are stronger than a single before-and-after comparison.

Evidence

Post-dismissal rebound

• What Makes Great Leaders & How to Live an Extraordinary Life | Dr. Jared Diamond says organizations often replace coaches or CEOs after unusually poor years and then credit them for improvement that may partly reflect regression toward average performance.

Partial leader contribution

• What Makes Great Leaders & How to Live an Extraordinary Life | Dr. Jared Diamond reports that coaches and CEOs account for meaningful but limited portions of outcome variation, requiring evaluation alongside players, company characteristics, industries, and institutions.

Counterevidence & Qualifications

Regression to the mean is not evidence that a replacement had no effect. A leadership change can alter selection, incentives, strategy, morale, practice, capital allocation, or information flow. The source provides conversational estimates rather than the underlying studies, so the concept supports a diagnostic question, not a universal numerical correction.

What Changed

• Created the concept as a causal-inference boundary around leadership replacement.
• Distinguished expected rebound from evidence of a specific new-leader mechanism.

Related Concepts

• Leadership Outcome Causation - broader model in which leaders contribute alongside people, institutions, environment, and chance.
• CEO Pay Luck Skimming / CEO薪酬揩油 - adjacent corporate-governance problem where external fortune is mistaken for managerial contribution.
• 军功归属错置 / Command Credit Misattribution - related error in assigning visible outcomes to the wrong actor or action.
• Good Company Vs Good Stock - separates organizational quality from outcomes driven by price and market conditions.
• Historical Counterfactual Reasoning - supplies the missing comparison logic for asking what would likely have happened without replacement.