Why Experience Doesn’t Protect Decision Quality

 A senior leader revisits a decision made several weeks earlier.

At the time, it felt efficient. The discussion was brief. Alignment came quickly. There was confidence, not hesitation. Yet the outcome now appears misaligned with how the situation actually evolved.

Nothing about the process felt careless.
If anything, it felt experienced.

This pattern is common among highly competent professionals - and it is often misunderstood.

Decision quality can degrade before anything feels wrong

Decision failures are usually analysed after outcomes deteriorate. Explanations then focus on missing information, flawed judgment, or external disruption.

What is less visible is that decision quality can degrade well before outcomes reveal a problem — often without producing a conscious sense of confusion, stress, or error.

Early changes tend to occur in how information is:

• sampled,

• weighted,

• and revised.

Decisions may feel clearer, faster, and easier to close.
Experience does not prevent this shift. In some cases, it obscures it.

What is decision quality degradation?

Decision quality degradation refers to a decline in how information is sampled, evaluated, and updated over time, even when intelligence, experience, and intent remain intact.

It typically occurs under changing operating conditions and often precedes visible decision failure. Importantly, it is not immediately experienced as poor judgment. It is experienced as efficiency.

Decision quality degradation is not the same as:

• lack of expertise,

• emotional decision-making,

• or irrational behaviour.

It reflects a change in how a cognitive system operates under load.

How experience changes tolerance for discrepancy

As expertise develops, professionals become better at recognising patterns and filtering noise. This is one of experience’s core advantages.

In decision science, this is often described using tolerance thresholds — the degree of discrepancy or disagreement required before a judgment is reconsidered.

With experience, tolerance thresholds tend to rise. Experienced decision-makers are less likely to revise established views in response to small or incremental deviations, particularly when confidence in prior judgments is high (Sun et al., 2024; Chen et al., 2024).

In stable environments, higher thresholds are functional. They prevent overreaction.

In dynamic environments, they delay adaptation.

Weak or early signals are filtered out. Discrepancies accumulate quietly. By the time thresholds are exceeded, decisions are often already committed.

When efficiency becomes entrenchment

Expertise also increases reliance on familiar patterns, routines, and response scripts — a phenomenon often described as cognitive entrenchment.

Entrenchment reduces cognitive effort and speeds execution. It also reduces sensitivity to novel conditions.

Research shows that experienced decision-makers may become less responsive to information that does not fit established schemas, particularly in unfamiliar or shifting contexts (Zhang, Harrington, & Sherf, 2022; Dai et al., 2026).

From the inside, the decision still feels coherent.
From the outside, it may be poorly aligned with emerging conditions.

Why confidence can suppress correction

Experience strengthens intuition. Intuition, however, is rarely examined when it feels reliable.

Evidence indicates that experts remain susceptible to cognitive biases and may be particularly vulnerable to unexamined overconfidence, especially in high-responsibility roles (Sherry et al., 2019; Pennycook, 2023).

As confidence increases, thresholds for detecting conflict or error also rise. Competing interpretations are dismissed earlier. Alternative perspectives are integrated less readily.

This does not feel like error.
It feels like clarity.

The experience paradox

These dynamics converge in what is often referred to as the experience paradox: greater experience does not reliably produce better decisions and may, under certain conditions, impair them.

Experience improves decision quality primarily when environments provide:

• high-validity cues, and

• timely, unambiguous feedback.

When feedback is delayed, incomplete, or ambiguous — as is common in leadership, strategy, and organisational contexts — experience alone does not improve decision accuracy (Liska, 2015; Peringa et al., 2026).

In such conditions, accumulated expertise can harden into earned dogmatism: justified confidence that gradually reduces openness to alternative interpretations (Chaney, Trelohan, & Moroz, 2025).

This is how experienced professionals are often blindsided.

When experience helps — and when it quietly hurts

Experience tends to improve decision quality when:

• environments are stable,

• patterns repeat,

• and feedback is fast and reliable.

Experience can degrade decision quality when:

• conditions change without clear signals,

• feedback is delayed or indirect,

• or decisions accumulate under sustained load.

The difference is not competence.
It is operating context.

Operating conditions that amplify degradation

The patterns described above are amplified under conditions of cognitive load.

Load does not require conscious stress. It can arise from:

• sustained responsibility,

• time pressure,

• information density,

• uncertainty,

• or prolonged decision exposure.

Under load:

• attention becomes more selective,

• heuristic reliance increases,

• tolerance for ambiguity decreases,

• and revision thresholds rise.

These responses are adaptive from a biological perspective. They trade flexibility for efficiency.

Selective attention and heuristic processing improve performance in familiar contexts but increase bias and oversight when conditions shift (Tejeiro et al., 2024). Ambiguity aversion under load can further delay revision or suppress exploration (Almazrouei et al., 2025).

Why decision failures feel unexpected

Decision degradation rarely announces itself.

It is experienced as:

• discussions closing faster,

• alternatives feeling less relevant,

• decisions repeating with minor variations.

By the time outcomes deteriorate, the internal decision system has often been operating under constrained conditions for some time.

Experience does not prevent this process.
It often masks it.

A systems view of decision quality

From a systems perspective, decisions do not fail because effort declines or competence disappears.

They fail because operating conditions change - and the cognitive system adapts in ways that prioritise efficiency over flexibility.

Experience raises tolerance thresholds.
Load increases rigidity.
Together, they can quietly degrade decision quality.

Understanding this pattern is not about correction or optimisation.
It is about recognising when decision quality starts to change — before consequences force recognition.

This pattern sits within a broader set of mechanisms that explain why capable people make systematically poorer decisions under load, explored in: Why Smart People Make Bad Decisions Under Stress.

References

Almazrouei, M., Dror, I., Morgan, R., Dan, O., Paterson, M., & Levy, I. (2025). Human factors in triaging forensic items: Casework pressures and ambiguity aversion. Science & Justice, 65, 102–111. https://doi.org/10.1016/j.scijus.2025.02.001

Chaney, D., Trelohan, M., & Moroz, D. (2025). When thinking you’re good makes you dumber: An investigation of consumers’ earned dogmatism. Journal of Business Research, 194, 115351. https://doi.org/10.1016/j.jbusres.2025.115351

Chen, H., Shao, L., Zhou, L., & Liu, J. (2024). Maximum consensus model with individual tolerance and mixed DEA prospect cross-efficiency for multi-attribute group decision-making. Applied Soft Computing, 158, 111572. https://doi.org/10.1016/j.asoc.2024.111572

Dai, F., Dong, L., Huang, L., & Tu, Y. (2026). The impact of ambidexterity on bid-winning performance: Evidence from an online crowdsourcing platform. Information & Management, 63, 104305. https://doi.org/10.1016/j.im.2026.104305

Lagziel, D., & Tsodikovich, Y. (2025). Working with AI: An analysis for rational integration. Games and Economic Behavior, 153, 241–257. https://doi.org/10.1016/j.geb.2025.06.009

Liska, A. (2015). What is intelligence? In Building an intelligence-led security program. https://doi.org/10.1016/B978-0-12-802145-3.00002-8

Pennycook, G. (2023). A framework for understanding reasoning errors: From fake news to climate change and beyond. Advances in Experimental Social Psychology, 67, 1–62. https://doi.org/10.1016/bs.aesp.2022.11.003

Peringa, I., Niessen, A., Meijer, R., & den Hartigh, R. (2026). Why experience fails to foster expertise in athlete selection. Psychology of Sport and Exercise, 82, 103022. https://doi.org/10.1016/j.psychsport.2025.103022

Sherry, J., Neale, T., McGee, T., & Sharpe, M. (2019). Rethinking the maps: Knowledge incorporation in Canadian wildfire risk management. Journal of Environmental Management, 234, 467–476. https://doi.org/10.1016/j.jenvman.2018.12.116

Sun, Q., Wu, J., Chiclana, F., & Ji, F. (2024). A tolerance index-based non-cooperative behaviour managing method with minimum cost in social network group decision-making. Expert Systems with Applications, 255, 124585. https://doi.org/10.1016/j.eswa.2024.124585

Tejeiro, R., Romero-Moreno, A., Paramio, A., Cruces-Montes, S., Galán-Artímez, M., & Santos-Marroquín, J. (2099). Maximization delays decision-making in acute care nursing. Scientific Reports. https://doi.org/10.1038/s41598-024-56037-x

Zhang, T., Harrington, K., & Sherf, E. (2022). The errors of experts: When expertise hinders effective provision and seeking of advice and feedback. Current Opinion in Psychology, 43, 52–57. https://doi.org/10.1016/j.copsyc.2021.06.011