Bolt 20 — State Is Not Ability

Bolt Series · Human Performance Calibration · Article 20

The student knew it yesterday. Why can’t they do it today?

This is one of the most frustrating experiences in education.

A child solves the problem perfectly during practice.

The next day, under examination conditions, the same child seems unable to begin.

What happened?

One possibility is that the original learning was never secure.

Another is that the capability exists but the current state is interfering with its expression.

Those are very different problems.

And they should not receive the same intervention.

Sport makes the difference visible

An athlete’s long-term ability does not guarantee identical output every day.

Sleep loss can impair performance.

Fatigue can impair performance.

Time of day can influence performance.

Recovery state can change performance.

A 2024 meta-analysis found that acute sleep deprivation negatively affected overall athletic performance, including speed, skill control, endurance and explosive performance.

A separate systematic review found measurable within-day variation in athletic output related to factors including body temperature and chronotype.

Those athletes did not become different people every few hours.

The underlying system was being expressed through different states.

Education has state effects too

Students also perform through changing states.

Sleep matters.

Stress matters.

Anxiety matters.

Attention matters.

Mental fatigue matters.

A 2023 systematic review and meta-analysis covering more than 53,000 primary-school children found test anxiety was negatively associated with Mathematics and literacy achievement and also related strongly to academic self-concept and self-efficacy.

A large 2026 meta-analysis also found a modest positive association between sleep quality and academic performance across school-age samples.

These findings do not allow us to explain every poor test result by anxiety or sleep.

They do tell us that observed performance is produced by more than stored knowledge alone.

A state can suppress a capability

Imagine a student can reliably solve simultaneous equations in normal practice.

During an examination, time pressure rises.

The student begins checking every line repeatedly.

Working memory becomes crowded.

They abandon a correct method midway.

The final answer is wrong.

What should we teach?

If we incorrectly diagnose “does not understand simultaneous equations,” we may reteach content the student already knows.

If the actual limiting factor is examination-state regulation, the repair belongs elsewhere.

But “state” must not become an excuse

There is a danger in this idea.

Every disappointing score could be explained away.

“I was tired.”

“I was stressed.”

“It was just a bad day.”

That would destroy calibration.

The answer is repeated evidence.

If the same weakness appears under many states, contexts and tasks, it is increasingly difficult to attribute it to one temporary condition.

If the weakness appears only under a particular condition and disappears reliably when that condition changes, that is meaningful too.

State is a hypothesis to test, not an excuse to assume.

This is why repeated measurement matters

A single score mixes together too many things.

Repeated performance lets us begin separating them.

Does the learner succeed without time pressure?
Does the same skill survive a new question format?
Does performance recover after rest?
Does the learner fail in the same place every time?
Does anxiety rise specifically during testing?
Is the apparent weakness stable across weeks?

The pattern begins telling us whether we are dealing primarily with capability, state, task design or some interaction among them.

State also changes the learner’s self-estimate

There is another subtle problem.

When people feel bad, they may judge their ability differently.

A tired student can interpret ordinary difficulty as evidence of incompetence.

An anxious student can mistake uncertainty for inability.

A highly energised student can mistake fluency for mastery.

So state affects both performance and the internal interpretation of performance.

That makes calibration more difficult—and more necessary.

A better language for one bad day

Instead of:

“I’m bad at this.”

try:

“My performance was poor today. I need to find out whether that reflects the skill itself or the state I performed it in.”

That statement does not protect the ego from evidence.

It protects the investigation from a premature conclusion.

Why this matters for parents and teachers

A child should not be permanently labelled from a temporary state.

But neither should important weaknesses be dismissed whenever conditions are imperfect.

The correct response is higher-resolution evidence.

Look across time.

Look across conditions.

Look across tasks.

Then update the model.

Performance is what happened. Ability is what we are trying to infer. State is one of the things standing between them.

Once we understand that difference, a surprising score becomes less of a verdict and more of a question worth investigating.