Where does overconfidence collide?

Being more certain that your own estimate is right than the evidence allows, so the range you would bet on is far too narrow: asked for a 90 percent interval, people give one that contains the true value only half the time. Lichtenstein, Fischhoff, and Phillips's 1982 Calibration of probabilities reviewed how stated confidence outruns accuracy, and Moore and Healy's 2008 The trouble with overconfidence labelled this specific error overprecision. This referent is a mismatch between one person's stated certainty and their real hit rate, and it is not about ranking yourself against other people, about trading in a market, or about a firm's decision to enter or acquire.[1][2]

Unit

One person's stated confidence set against their actual accuracy

Recovery means

Feedback and scoring rules can tighten calibration, though the effect is stubborn

Measured by

By calibration curves and by how often stated confidence intervals actually contain the true value

Believing you rank above other people more than the facts warrant, as when most drivers rate themselves safer than average: the error is comparative, about your place in a group, not about your absolute accuracy. Moore and Healy's 2008 The trouble with overconfidence separated this better-than-average effect, which they called overplacement, from being too sure of your own estimate. This referent is a claim about relative standing among people, and it can even run the other way on hard tasks, so it is distinct from a too-tight private certainty, from an investor's trading, or from a manager's bid for a target.[2]

Unit

One person's judged rank against the true rank within a group

Recovery means

Not applicable

Measured by

By comparing self-rated percentile standing with measured performance across a group

Investors who overrate their private information and skill trade too often, and the costs of that churn leave them worse off than if they had held still. Odean's 1998 Volume, volatility, price, and profit when all traders are above average modelled how such traders inflate volume, and Barber and Odean's 2001 Boys will be boys showed with brokerage data that the most active accounts earned the least after costs. This referent is a market behaviour with a measurable price, trading turnover and net return, not a private calibration error, a comparative self-ranking, or a firm's strategic move.[3][4]

Unit

A trader's turnover and net return, read from account records

Recovery means

Not applicable

Measured by

By trading turnover and by risk-adjusted net returns after transaction costs

Managers and founders who overrate their own judgment enter crowded markets and pay too much to acquire other firms, so the confidence shows up as a costly organisational decision rather than a private belief. Camerer and Lovallo's 1999 Overconfidence and excess entry found experimentally that people crowd into competitive entry because each expects to be the winner, and Roll's 1986 The hubris hypothesis of corporate takeovers argued that acquirers overpay because they trust their own valuation over the market's. This referent is a firm-level strategic outcome, entry rates and acquisition premia, distinct from calibration, self-ranking, or an investor's turnover.[5][6]

Unit

A firm's entry choice or the premium it pays to acquire a target

Recovery means

Not applicable

Measured by

By rates of market entry and by the size of premia paid in acquisitions