Small Group Tutorials

Here to help students catch up, keep up, and move ahead. Book a consultation here.

Chat on WhatsApp

Bolt 08 — When You Think You Are Worse Than You Are

Bolt Series · Human Performance Calibration · Article 08

What if the evidence says you can do more than you think?

A student predicts 48%.

The paper returns at 84%.

Everybody is delighted.

But from a calibration point of view, something is still wrong.

The student’s performance was much stronger than the student expected.

We usually worry about the opposite problem: people thinking they are better than they are.

But thinking you are substantially worse than repeated evidence shows can also distort decisions.

Underestimation is not automatically humility.

Sometimes it is simply another form of miscalibration.

Research does find underconfidence

Metacognitive research has documented situations in which people underestimate later performance.

In classic learning experiments, judgements of learning sometimes became underconfident after repeated study-and-test cycles: actual performance improved more than participants’ estimates of what they would remember.

Other work has shown that this effect has boundary conditions. Underconfidence is not a universal law, and task difficulty, judgement timing and the cues people use can change the direction of the error.

In classroom research, high-performing students are sometimes found to underestimate their performance while lower-performing students more often overestimate. One introductory biology study found that practice testing was associated with underconfidence among high-performing students even as it improved calibration for many learners.

The pattern is not simple enough for stereotypes.

But the existence of underestimation is enough to matter.

Being capable and feeling capable are different things

Suppose a student can solve difficult algebra problems correctly eight times out of ten.

But before every attempt, the student thinks:

“I probably can’t do this.”

The underlying capability and the internal estimate are no longer aligned.

That matters because the estimate helps decide what happens next.

Do I attempt the harder problem?
Do I volunteer an answer?
Do I apply for the programme?
Do I sit the more demanding paper?
Do I trust a correct solution long enough to submit it?
Do I keep revising something already secure while neglecting something weaker?
Do I ask for help because I need it, or because I cannot recognise that I already know?

A low estimate can therefore reduce access to capability that genuinely exists.

This is not an argument for blind positivity

The solution is not to tell every uncertain student:

“You are amazing. Stop doubting yourself.”

That simply replaces one unsupported estimate with another.

The better response is evidence.

“You predicted around 50 on the last three papers. You scored 78, 82 and 84. Let’s look at why your estimate remains so much lower than your performance.”

Now the learner has something solid to inspect.

We are not asking the child to believe a compliment.

We are asking the child to update from receipts.

Past failure can become an outdated measuring instrument

There is a particularly interesting reason people can underestimate themselves.

Their internal estimate may still be using yesterday’s version of themselves.

A child struggled badly with fractions two years ago.

They practised.

They improved.

Current evidence is strong.

But when a fraction appears, the old internal message arrives first:

“I’m bad at fractions.”

The learner has changed.

The self-estimate has not.

Research on underconfidence with practice gives us a related laboratory example: after repeated learning, people can sometimes base later judgements too heavily on earlier performance and fail to fully credit improvement that has already occurred.

This gives teachers a powerful question:

“Are you judging what you can do now, or what you could do before?”

Strong students can waste effort too

Underestimation can look responsible.

The student studies everything again.

Checks every answer five times.

Refuses to move on until certainty feels perfect.

But time is finite.

If a learner cannot recognise what is already secure, they may spend scarce revision time protecting strengths while leaving true weaknesses untouched.

Calibration therefore affects resource allocation in both directions.

Overestimation can make us stop too early.

Underestimation can make us continue too long.

Both can be inefficient.

Sport helps because performance is repeatedly visible

An athlete can feel uncertain about readiness.

Then training and competition provide repeated evidence.

The point is not that elite athletes always know themselves perfectly. They do not.

The useful feature of sport is that estimates can be tested often.

Times.

Distances.

Loads.

Video.

Competition.

And subjective reports matter too. Sports science routinely uses athlete-reported ratings of perceived exertion because internal experience can contain information that external measurements do not fully capture. At the same time, research warns that subjective measures require careful validation and should not automatically be treated as perfect readings.

Again, the solution is not to choose one observer.

It is to compare readings.

A practical classroom correction

If a student repeatedly underpredicts, do not merely praise them after each unexpectedly good result.

Make the pattern visible.

Prediction: 60. Result: 79.
Prediction: 55. Result: 81.
Prediction: 65. Result: 84.

Then ask:

Why do you keep predicting lower?
Which evidence are you ignoring?
What would a more reasonable prediction have been?
Which parts are genuinely uncertain and which are now secure?
What new challenge would be appropriate if this performance is real?

The goal is not to manufacture confidence.

It is to correct the estimate.

Why this matters for education

A student who persistently underestimates themselves may carry real capability that they do not use when decisions have to be made.

They may choose the easier path, remain silent, seek unnecessary reassurance, avoid unfamiliar work or refuse opportunities that their actual performance suggests they can handle.

That does not mean every hesitant child is secretly exceptional.

It means the same rule applies in both directions:

Your estimate should move when repeated evidence says it is wrong.

If reality repeatedly says you need more work, update downward and work.

If reality repeatedly says you can already do more than you think, update upward and act accordingly.

Calibration is not pessimism.

It is not optimism.

It is the discipline of allowing evidence to change what you believe about your current capability.