Here is what you might expect: the harder the task, the less confident you feel about pulling it off. The easier the task, the more assured you are.

Here is what the evidence shows: it tends to go the other way. People systematically overestimate their chances on tasks described as difficult, and underestimate their chances on tasks described as easy. Confidence runs backwards relative to actual performance probability.

This is the hard-easy effect, and it is one of the more counterintuitive findings in the study of self-assessment.

What it is

The hard-easy effect is the tendency to overestimate one's ability to succeed at difficult tasks and underestimate one's ability to succeed at easy ones. The miscalibration is direction-specific - it is not simply that people are overconfident or underconfident in general. The error tracks the difficulty of the task, and tracks it in the wrong direction.

In experiments, participants asked to predict their performance on a range of tasks show a consistent pattern: they rate their chances on hard tasks higher than their actual success rate, and rate their chances on easy tasks lower than their actual success rate. The harder the task, the more inflated the confidence; the easier the task, the more deflated.

The mechanism: why difficulty distorts the signal

The hard-easy effect arises from how difficulty cues interact with heuristic self-assessment.

When a task is presented as hard, it activates a sense of challenge - and challenge activates a sense of personal agency. People think about effort, determination, and their own drive to succeed. These are not assessments of objective probability; they are self-representations. The result is an inflated estimate of personal performance capacity that exceeds what the base rate of success on such tasks would predict.

When a task is presented as easy, the opposite framing takes hold. Easy tasks feel trivial. The person mentally sets the task aside as already handled, or assumes success is so certain that their personal skill is irrelevant - which paradoxically reduces the sense of personal contribution and, with it, the confidence estimate. Something like: if anyone could do this, my particular ability to do it is nothing special.

Both distortions come from the same source: the difficulty label is being processed as information about self-efficacy rather than as information about objective task complexity. The label hijacks the self-assessment rather than informing it.

The myth and the evidence

The myth: being told a task is hard makes people appropriately cautious; being told it is easy makes them appropriately relaxed.

The evidence: being told a task is hard makes people inflate their self-assessment; being told it is easy makes them deflate it.

This matters practically because it means that the natural emotional response to difficulty - the arousal of personal challenge - actually degrades the accuracy of self-prediction. Feeling motivated and challenged by a difficult task is a useful drive. But the associated confidence boost is not calibrated to reality; it is a product of the framing, not the skill.

Where it shows up

In planning. Someone taking on a genuinely difficult project - a career change, an ambitious renovation, a complex negotiation - may underestimate the actual probability of running over time, over budget, or into unforeseen obstacles because the challenge activates an inflated sense of capability. Meanwhile, the routine tasks around the project - the straightforward communications, the familiar administrative work - get underplanned because they seem too simple to bother about.

In interpersonal judgments. When someone we care about takes on something difficult, we often tell them - and ourselves - that they will be fine. The difficulty of the thing registers as evidence of their determination rather than as a reason for realistic concern. When they take on something easy, we sometimes underestimate their chances, dismissing it as too basic to be reliable. Both reads can be wrong.

In performance feedback. If a manager evaluates an employee's performance on a tough assignment as strong and their performance on a routine task as weak, it is worth asking whether the hard-easy effect is running. The manager's confidence calibration may be tracking difficulty, not performance quality.

What the hard-easy effect is not

It is not the same as general overconfidence. General overconfidence is a tendency to be too confident across the board. The hard-easy effect is direction-specific - the miscalibration depends on the difficulty of the task being assessed. The two can co-occur, but they are distinct patterns.

It is not the Dunning-Kruger effect. The Dunning-Kruger effect concerns the relationship between actual skill and metacognitive awareness - lower skill sometimes correlating with less accurate self-assessment. The hard-easy effect concerns difficulty-based miscalibration and can appear across skill levels.

It does not only affect novices. Experienced professionals show the hard-easy effect too. The mechanism is heuristic thinking triggered by difficulty cues, not a lack of domain knowledge.

Sources