How can we learn to recover from failure?

TL;DR. Recovering from failure involves several abilities: making room for disappointment, examining the facts and preparing a different attempt. Research offers ways to work on each of these, without guaranteeing success next time. This reflection begins with what I hope to pass on to my daughter and ends with a practical guide to practising, asking for help and deciding when to persist or change direction.

Wooden blocks scattered and assembled on a workbench.

Preparing another attempt using what the previous one helped us understand.

Among the things I would like to pass on to my daughter is the ability to recover from failure. Being able to work through disappointment, understand mistakes and find a way forward seems useful in many situations in life.

I started with my own experience, then wanted to explore further. I looked at what scientific research, books and accounts of people’s lives could tell us about this ability. How do we learn it? What helps us try again? And how do we distinguish persistence from repetition that wears us down?

This article is a reflection on what I take from that work. It brings together ideas I would like to pass on, followed by a practical way to put them to use.

Contents

What does recovery mean?
How can we examine our mistakes without condemning ourselves?
When does reflection become useful?
What can we learn from Darwin and Tu Youyou?
How can we prepare a different attempt?
When is it time to change direction?
What I hope to pass on
A practical guide: preparing to try again after failure
Further reading

What does recovery mean?

We can start working on a project again while still feeling disappointed. We can also feel better without understanding why the first attempt failed. Emotional recovery, taking action again and learning sometimes move at different speeds.

This distinction interests me because it leaves room for imperfect returns. Waiting until all our confidence has come back can delay a step that is already within reach. Conversely, getting back to work quickly does not, by itself, show that we have recovered or learned something.

There is also a difference between learning from failure and succeeding next time. Our preparation can improve without the desired result arriving immediately. Some circumstances remain uncertain, some decisions belong to other people, and the resources available matter.

Research on resilience describes it as a process that also depends on relationships and the environment. Having someone to talk to, time to recover or a financial cushion changes the possibilities for trying again. These conditions belong in the explanation alongside individual effort. Southwick, Charney and DePierro, Resilience, 2023, excerpt from the first chapter.

For me, learning to recover therefore includes recognising what we need. Asking for specific feedback or seeking practical support may be the most useful next step.

How can we examine our mistakes without condemning ourselves?

“That presentation went badly” describes an event. “I’m incapable of speaking in public” draws a much broader conclusion. Between the two, information about what happened can disappear: preparation time, the structure, the audience’s questions and the parts that worked.

Self-compassion, as studied in psychology, involves approaching our difficulties with less hostility towards ourselves. This may seem at odds with holding ourselves to high standards. Yet a series of four experiments by Juliana Breines and Serena Chen suggests that this attitude can support the motivation to improve.

In one experiment, after struggling with a test, participants encouraged to take a compassionate attitude towards themselves spent more time studying than those in the no-intervention group. There was no demonstrated improvement in their next test score. The finding concerns preparation effort, which is useful information as long as we keep that distinction. Breines and Chen, 2012.

I take from this the possibility of being more precise in what we ask of ourselves. We can acknowledge a mistake, accept its consequences and work on it without making it a definition of who we are. A sentence such as “I’m disappointed with this result, and I can examine what led to it” leaves that possibility open.

This does not require convincing ourselves that everything is fine. Disappointment has its place. The aim is to regain enough mental space to notice what can be corrected.

When does reflection become useful?

Returning to an event after a setback can help us understand it. We can also spend a long time returning to the same accusations. Time spent thinking does not, on its own, tell us whether our understanding is improving.

Edward Watkins’s work distinguishes different forms of repetitive thought. Their effects depend partly on context and on the level at which we think. Concrete reflection on how a situation unfolded can be more useful than a general evaluation of ourselves. This does not make every question beginning with “why” unhelpful; it invites us to notice what the question actually helps clarify. Watkins, 2008.

In the presentation example, “Why am I so bad at this?” gives us little to work with. “At what point did I lose my train of thought?” helps identify a moment. We can then examine whether the outline was too crowded, whether a question interrupted our reasoning or whether we had mostly rehearsed by reading our notes.

This resembles a debrief: comparing what was planned with what happened, then deciding on an adjustment. A meta-analysis of 61 studies found benefits of debriefing on the training outcomes examined. Objective observations and matching the debrief to the work being reviewed matter to these results. Keiser and Arthur, 2021.

This is also part of how we organise learning, a question I explore in this French-language essay: providing opportunities to practise, followed by feedback that allows something to change. “Lacking confidence” sometimes describes how we feel. We still need to identify what would help us do better next time.

What can we learn from Darwin and Tu Youyou?

Biographies show how people worked with uncertainty. For this question, their value lies in the actions they make visible. A successful life does not prove that the same method would have produced the same outcome for everyone.

In his autobiography, Charles Darwin explains that he recorded facts and observations that contradicted his ideas. He had noticed that he forgot these more readily than those that supported them. This detail seems particularly useful after a failure: keeping the information that might require us to revise our explanation. Darwin, autobiography published in 1887.

If I explain a difficult presentation solely by saying that the audience was unreceptive, I may miss a problem with clarity. If I explain it solely in terms of my own inability, I may miss the fact that some people understood very well. Writing down what challenges my initial account keeps several explanations in play.

Tu Youyou describes another action in her autobiographical account for the Nobel Prize. Early sweet wormwood extracts tested against malaria produced variable results. An ancient text helped direct her attention towards the extraction conditions, particularly temperature. Her team changed the procedures and continued testing. Tu Youyou, Nobel autobiography, 2015.

The ancient text provided a lead; experiments were used to assess it. This account reveals collective work, technical choices and changes in method. Persistence takes a concrete form here: looking for what deserves to change in the previous attempt.

These accounts help me formulate a more useful question than “Have I tried hard enough?”: what will the next attempt help us understand that the previous one did not clarify?

How can we prepare a different attempt?

Amy Edmondson offers criteria for recognising failures that can genuinely inform exploration. These include new territory, a worthwhile goal, an informed hypothesis and an experiment kept as small as possible to learn what is needed. This framework comes from her work on organisations. Edmondson, Harvard Business School interview.

Applied to an everyday situation, this points towards reducing the size of the next attempt. Before delivering an entire presentation again, we could try the introduction with one person and ask them to restate the main idea. The feedback becomes easier to interpret, and the attempt takes less time.

The choice of what to change matters as much as its size. If we change the topic, the structure and the way we rehearse all at once, an improvement will remain difficult to explain. Where possible, isolating one change helps us understand what made a difference.

Research on error management training provides additional support. A meta-analysis of 24 studies found benefits from training that allows people to explore, make errors and learn to manage them. These findings concern training situations designed for that purpose, where the consequences of errors are kept under control. Keith and Frese, 2008.

Then comes the task of starting. An intention such as “I’ll prepare better” leaves many decisions to be made at the moment of action. A plan can specify the expected obstacle and a response: “If I feel like postponing practice because I’m afraid of the result, then I’ll rehearse only the introduction, at the planned time.”

Research combining a desired outcome, a concrete obstacle and an “if… then…” plan finds a positive average effect on goal attainment. The effect remains modest and varies across situations. A plan helps organise action; it does not replace the resources needed to carry it out. Wang et al., 2021 meta-analysis.

This attention to the step between intention and action also appears in my reflection on Pokémon and Health: making an action accessible, allowing it to recur in everyday life and making progress visible.

When is it time to change direction?

The ability to try again includes a sometimes difficult decision: determining what we still want to pursue. A goal can remain meaningful while the method needs to change. It can also lose its importance or require resources we would rather devote elsewhere.

Research on goal adjustment distinguishes several processes: disengaging from a goal that has become unattainable, reengaging elsewhere or remaining flexible in how we move forward. A large meta-analysis published in 2025 examines their relationships with well-being and functioning. Much of the evidence remains associational; it does not provide a universal rule for deciding when to stop. Riddell et al., 2025.

In making this decision, I find it useful to separate the value of the goal from attachment to the attempt already underway. Learning to communicate an idea may still matter, even if we choose a format other than public speaking. Equally, another presentation may be a reasonable attempt if specific feedback has helped us prepare a change.

Criteria set before trying again make the decision clearer: the time we are willing to spend, the help available, what we hope to observe and the information that would lead us to revise the project. A pause can then have both a reason and a condition for resuming.

What I hope to pass on

I return to the thought of my daughter with a clearer idea of what I would like to pass on. Being able to say that an attempt was difficult. Finding someone to look at it with. Understanding that a mistake can be examined without calling the whole person into question.

I would also like to pass on the freedom to change methods, and the possibility of asking for help before knowing exactly how to begin again. In this reflection, confidence seems like something we can rebuild as we go, with support and opportunities to learn.

The studies cited largely concern adults or specific tasks. What I take from them for my own child reflects my choices as a parent; it is not a validated educational programme. The guide below is for someone who wants to work on this ability for themselves, starting from a concrete situation.

A practical guide: preparing to try again after failure

This outline brings together the ideas above. It has not been tested as a complete protocol. To try it, choose a setback whose consequences remain manageable: a disappointing presentation, a failed exam or an approach that received no response. Timing and pace should fit the situation.

1. Regain enough mental space to examine the facts

Start by naming how the event makes you feel and what would help you look at it. That might mean a conversation, some rest or a planned time to come back to it. The aim is to examine the situation with a little distance, even if disappointment is still present.

One possible formulation: “I’m disappointed with this result. I’ll be better able to understand it after resting and asking for feedback.”

2. Describe a specific event

Write down what you expected, what happened and the evidence available. Set aside general conclusions about yourself, and also record what worked.

In our fictional example: “I wanted to present an idea in ten minutes. I took fifteen. The person listening understood the problem but did not identify my proposal.” This description gives us specific points to examine.

3. Look for explanations you can check

Briefly reconstruct what happened and write down two plausible explanations. Distinguish what depended on preparation, what depended on the context and what remains unknown. For each explanation, note what could support it or call it into question.

For example: “My proposal comes too late in the outline” and “I assume people know terms they are unfamiliar with.” Then ask for specific feedback: “At what point did the proposal become clear? Which words needed an explanation?”

4. Choose a limited attempt and get the help you need

Decide on one main change and a useful observation. Check that the resources are available: someone to give feedback, a time slot, equipment or additional learning.

For example, state the proposal at the beginning, then deliver that introduction to someone unfamiliar with the project. Ask them to restate what they understood, without completing their answer for them. Limiting the attempt to this passage makes it possible to work on clarity before repeating the whole presentation.

5. Plan the start and the next decision

Choose a realistic time and write an “if… then…” plan for the most likely obstacle. Also decide in advance what would lead you to continue, modify or pause the attempt.

For example: “On Wednesday, after lunch, I’ll rehearse the introduction. If I start rewriting the entire presentation to put that moment off, then I’ll return to this passage and record it.” If the proposal is understood, the next step could be working on the overall length. If it remains unclear, you could examine the wording with the person giving feedback.

6. Review what the attempt taught you

After the attempt, compare the observation with what you expected. One specific element may have improved while the overall result remains unsatisfactory. No improvement may be a reason to test the second explanation or ask for another perspective.

To develop this ability over time, use the same approach with other manageable situations. Notice whether your analysis becomes more precise, whether the help you request becomes more useful and whether subsequent attempts actually change. No set number of days guarantees that you will acquire this ability.

A worksheet to keep

What to fill in	Fictional example
What happened	Presentation too long; proposal not clearly understood.
How I feel and what would help	Disappointed; need rest and specific feedback.
What worked	The initial problem was understood.
Explanations to check	Proposal introduced too late; unfamiliar vocabulary.
The next attempt	State the proposal in the introduction.
The support needed	Someone unfamiliar with the project.
My plan for the obstacle	If I postpone by rewriting everything, record only the introduction.
The observation and next decision	If the proposal is understood, work on length; otherwise, examine the wording.

The worksheet can stay short. Its purpose is to keep a clear enough record to return to the work, including when confidence takes longer to return.