Aviation Safety Safe Fly Aviation · August 12, 2026 · 13 min read

Pilot Decision-Making Under Pressure: Why Experience, Proficiency and Check Rides Are Not Enough

Written by Sudip Sharma, Director, Safe Fly Aviation Services Pvt. Ltd.

Why do experienced, qualified and technically proficient pilots sometimes make decisions that appear obviously wrong in hindsight? The answer has less to do with skill than with how good decisions quietly become bad ones under pressure.

Pilot Decision-Making Under Pressure

We’ve all been there. Or at least, we’ve all trained for that moment in the simulator. The instructor throws a curveball—an engine failure on takeoff, a sudden weather diversion, a hydraulic warning light that shouldn’t be on. We run the memory items, call out the QRH procedures, and land safely. The check airman nods, signs the paperwork, and we walk away with another tick in the box.

That is not the hardest question in aviation safety.

The harder question is this: why do experienced, well-trained, technically proficient pilots sometimes make decisions that look obviously wrong in hindsight? Not decisions made by novices who lacked the skill to do better—decisions made by crews who had the skill, the qualifications, and the experience, and still chose the wrong course of action.

This is not a story about incompetence. It is a story about how good decision-making erodes under pressure, even in cockpits staffed by highly capable people. Understanding that erosion—and building the training, culture, and habits that resist it—is the real work of aviation safety.

Qualified Does Not Always Mean Competent

Passing examinations, simulator checks, and type-rating tests creates a legitimate sense of readiness. These events are designed, observed, and graded. They follow a predictable script. The failures are known in advance. The check airman is expecting certain errors, and knows what to look for.

Real operations are not like that.

The weather doesn’t follow the forecast. Maintenance issues don’t schedule themselves for convenient times. Controllers don’t always hand you the approach you planned. And your body doesn’t care that you need to be sharp—fatigue is fatigue, and it erodes judgement long before it erodes technical skill.

The gap opens when crews treat a successful check ride as a permanent certificate of operational judgement, rather than a snapshot of technical performance on a particular day.

A proficiency check demonstrates performance against defined regulatory and operator standards. It does not, by itself, provide a complete picture of how a crew will make decisions in every complex operational situation—correctly interpreting a subtle automation mode change, rejecting an unstable approach late in the sequence, or diverting when the destination weather is still technically legal but operationally marginal.

This gap is not a failure of individual pilots. It is a structural feature of how we have traditionally measured readiness. We test what we can easily measure—procedure compliance, manoeuvre standards, systems knowledge—and assume the harder-to-measure skills will follow. They don’t always. And when they don’t, the consequences can be catastrophic.

Proficiency vs Competence

Proficiency is generally demonstrated against defined performance standards—regulatory, type-specific, or operator-specific. It answers a straightforward question: “Can the pilot perform the required procedures to the required standard?” It’s the ILS to minimums. It’s the single-engine go-around. It’s the stall recovery. It’s the memory items for the engine fire. You can test proficiency in a simulator on a Tuesday morning with a fresh cup of coffee and a check airman following a script.

Competence is the broader ability to apply knowledge, skills and judgement effectively in an operational context—including, but not limited to, the technical procedures a proficiency check measures. It answers a more demanding question: “Can the pilot recognise what is happening, evaluate the options, and choose the course of action that best protects the flight when the situation is incomplete, time-pressured, or conflicting?”

Well-designed competency frameworks generally treat technical and non-technical skills together, rather than as a simple proficiency-versus-competence split. The distinction that matters operationally is less about which skills are involved, and more about which conditions are being tested: a scheduled check against a known scenario, versus a live situation with incomplete information and no script.

A fully competent pilot integrates several interdependent skills that go far beyond raw stick-and-rudder ability:

• Solid technical knowledge of systems, performance, and limitations
• Precise aircraft handling, both with and without automation
• Operational judgement that weighs weather, fuel, terrain, traffic, and passenger needs against risk
• Decision-making under uncertainty
• Continuous situational awareness
• Clear communication and effective teamwork
• Honest recognition of personal and crew limitations

Proficiency can be demonstrated in a simulator on a scheduled check day. Competence—and the decision-making it produces—is revealed on the line, at 0200 in deteriorating weather, with incomplete information and a mission that still feels achievable.

Why Experienced Crews Can Still Make Poor Decisions

It’s tempting to assume experience is protective—that the more hours a pilot has flown, the less likely they are to make a poor decision. Experience does improve judgement, on average. But it is not a guarantee, and in some circumstances it can work against a crew.

Experienced pilots have seen more edge cases, which builds real, valuable pattern recognition. But that same familiarity can, in some circumstances, contribute to:

• Overconfidence — a long track record of successful outcomes can create a quiet sense that this situation, too, will resolve itself, even when the objective indicators say otherwise.
• Familiarity with operating margins — experienced pilots often know how much margin an aircraft, a procedure, or a limit actually has in practice, and can be tempted to use more of it than a newer pilot would.
• Normalisation of deviation — if a marginal practice has worked without incident ten times before, it starts to feel like standard practice rather than a risk being carried forward.
• “I’ve seen this before” thinking — pattern-matching a new situation to a past one that resolved safely, even when the current situation is following a different, more dangerous trajectory.
• Resistance to changing the original plan — the more experience and authority a pilot has, the more psychologically costly it can feel to reverse a decision in front of a crew, passengers, or company.

None of this means experience is a liability. It means experience should improve judgement, but it does not, by itself, guarantee good judgement. The two have to be built and maintained together—through training, culture, and honest self-assessment—not assumed to travel as a package.

How Bad Decisions Develop

Accident investigators rarely find a single cause. What they find, again and again, is a chain—a sequence of small, individually defensible decisions that combine into an outcome no one would have chosen if they’d seen the whole picture at the start.

That chain tends to follow a recognisable shape:

1. A small threat appears. A weather cell forming near the destination, a system indication that’s probably nothing, a delay that eats into the fuel margin.
2. The threat is normalised. It’s noted, but it doesn’t change the plan. Everyone has seen worse.
3. The crew commits to a plan. A destination, an approach, a course of action—and begins to invest psychologically in seeing it through.
4. Alternatives quietly narrow. As time passes without the plan being revisited, options fall away—the divert airport closes, the fuel margin tightens, the daylight runs out.
5. Workload increases. The situation now demands more attention than the plan accounted for, leaving less capacity to step back and reassess.
6. Challenge weakens. Under load, callouts get softer, questions go unasked, and the monitoring pilot’s doubts stay unspoken.
7. Margins deteriorate. Speed, altitude, fuel, or terrain clearance erode further from the safe range—often without one dramatic moment marking the change.
8. The decision comes late. By the time it is made, it is being made under far worse conditions than if it had been made ten minutes earlier.
9. Recovery options are lost. The safest choices that existed earlier are no longer on the table.

None of these steps looks alarming in isolation. That is exactly what makes the chain dangerous—each link is a small, defensible decision, and the crew rarely experiences a single moment that feels like “the wrong choice.” The wrong choice is usually the accumulation of several reasonable-sounding ones.

The sections below examine the individual mechanisms—plan continuation, confirmation bias, authority gradient, automation dependence, and workload saturation—that drive this chain forward. Recognising them individually is useful. Recognising the chain they form together is what actually changes outcomes.

The Psychology Behind Plan Continuation

Plan continuation bias is the tendency to continue with an existing plan despite changing circumstances. In operational aviation, this can manifest as the familiar “get-there-itis” mindset—the powerful pull to press on once committed to an approach or a destination, even as conditions change.

This is chain steps three and four in action: the plan was made when conditions looked different, and every minute that passes without a formal decision to change course reinforces the original plan by default. Continuing starts to feel like the natural state; diverting starts to feel like the exceptional one—even when the data has quietly reversed which option is actually safer.

Confirmation Bias and “One More Attempt”

Confirmation bias is one of the most dangerous mechanisms in the chain. Once a crew has committed to a plan—we’re going to land at Destination X—they tend to seek out information that supports that plan and discount information that contradicts it. The ATIS says visibility is still above minimums? Good, we’re pressing on. The fact that the trend is rapidly deteriorating? That gets filed away as something to monitor, not something to act on.

This is where “one more attempt” thinking takes hold: a missed approach doesn’t trigger a divert decision, it triggers a second attempt, because the first attempt gets reframed as bad luck rather than as new information about a worsening trend. Each additional attempt consumes fuel, time, and margin—and makes the eventual decision to divert look, in the moment, more costly than it did on the first approach, even though it is now the safer choice by a wider margin.

Authority Gradient: When the First Officer Doesn’t Challenge

Authority gradient remains a persistent challenge, particularly in organisational cultures where hierarchy, seniority, or communication norms make challenge uncomfortable. The first officer sees something wrong. They know they should speak up. But the captain is senior, confident, and doesn’t seem to be looking for input. So they stay quiet, and step six of the chain—weakened challenge—takes hold exactly when it’s needed most.

A steep authority gradient doesn’t require a hostile captain. It can exist quietly, by default, in any cockpit where challenge has never been explicitly invited and rehearsed.

Automation and the Loss of Situational Awareness

Modern flight decks can shift a pilot’s role from actively flying the aircraft to monitoring a system that is flying it instead. That shift is not inherently a problem—automation reduces workload and, used well, improves precision. The risk is not the automation itself. The risk is the gap between what the system is doing, what the crew believes it is doing, and how quickly the crew recognises the difference.

When that gap opens, loss of situational awareness often follows directly. The crew goes heads-down to work out why the aircraft’s behaviour doesn’t match their expectation, and loses track of the larger picture—altitude, terrain, fuel, or sequencing—while they do it. By the time they look up, the situation has moved on, and the safe window to act has narrowed further.

Workload Saturation and Cognitive Overload

Workload saturation is the point at which additional information can no longer be processed effectively. The human brain has limited bandwidth, and when it’s full, it’s full. The pilot who is flying the aircraft, troubleshooting a system fault, coordinating with ATC, and monitoring deteriorating weather is simply not going to do all of those things well at once. One of the risks is that the crew loses the cognitive capacity needed to step back and reconsider the plan.

When the Checklist Cannot Make the Decision for You

Checklists are indispensable. They standardise critical actions and protect against memory lapses. They’re the backbone of disciplined operations. Yet a checklist cannot replace the operational judgement required to recognise when a situation demands a change of plan.

Competence determines the far more difficult judgements:

• When to stop the approach and go around
• When to divert rather than continue to the destination
• When to reject a landing after the wheels have touched down
• When to declare an emergency early rather than late
• When to challenge the other pilot—and how firmly

The most dangerous moments in aviation are often those for which no single checklist fully applies. The crew must synthesise incomplete data, time pressure, and competing goals. That synthesis is competence. It’s not something that can be captured in a one-size-fits-all procedure, because every situation is unique.

Consider the approach that’s stable on paper but feels wrong. The speed is on target. The glidepath is captured. The landing checklist is complete. But the wind shear alert just came on, the rain is picking up, and the captain has a nagging sense that something isn’t right. There may be no single checklist item for that combination of warning signs. The crew still has to interpret the overall risk.

There’s only judgement. And that judgement is what separates the competent pilot from the merely qualified one.

“A highly competent pilot is not the pilot who always completes the mission. It is the pilot who knows when the safest decision is to stop.”

The Last Safe Opportunity

If there is one habit that would prevent more accidents than any single piece of technology, it is this: for every developing situation, ask “What is our last safe opportunity to change the plan?”

Not the ideal opportunity. Not the comfortable opportunity. The last one that still leaves genuine options on the table.

That question applies constantly, not just at the obvious decision points:

• The last safe opportunity to discontinue an approach — before the aircraft reaches a point where the available recovery options become more constrained.
• The last safe opportunity to divert — before fuel, weather at the alternate, or daylight remove it as a realistic choice.
• The last safe opportunity to reduce workload — before task saturation makes it impossible to ask for help or reprioritise.
• The last safe opportunity to declare an emergency — before the situation deteriorates to the point where declaring it early would have bought meaningfully more support and priority.
• The last safe opportunity to disconnect automation — before an unexpected mode or behaviour has already taken the aircraft somewhere the crew didn’t intend.
• The last safe opportunity for the monitoring pilot to intervene — before a soft challenge that goes unanswered becomes the only challenge that was made.

Every one of these moments passes quietly. There is rarely an alarm that says “this is the last safe opportunity.” That is precisely why competent crews build the habit of asking the question proactively, on a schedule, rather than waiting for a trigger that may never announce itself clearly enough.

This is the practical answer to knowing when to say no: not waiting for certainty that something is wrong, but continually protecting the option to act while it still exists.

This entire pattern—recognising threats, weighing alternatives, and choosing the safest available action—is sometimes described formally as Aeronautical Decision Making (ADM): the practical application of judgement, risk assessment, and available information to operational decisions. CRM and TEM, covered next, are best understood as structured ways of teaching and reinforcing ADM within a crew environment, rather than as separate disciplines.

CRM: Creating a Cockpit Where Challenge Is Expected

Some pilots still roll their eyes when Crew Resource Management is mentioned. They see it as “touchy-feely” training, a box to be ticked during recurrent, not something that has any bearing on how they actually fly the aircraft.

That view is wrong. And more than that, it’s dangerous.

Modern CRM, as defined by ICAO, explicitly includes situational awareness, leadership, decision-making, stress management, and Threat and Error Management. It is not soft-skills training bolted onto the technical syllabus. It is the operating system that allows technical proficiency to be applied safely under real operational pressure.

Operators that treat CRM as a periodic classroom module miss its operational purpose. CRM only delivers value when it is assessed in realistic scenarios and reinforced every day on the line. It needs to be practised, not just taught. And it needs to be evaluated, not just assumed.

The captain who genuinely encourages challenge from the first officer. The first officer who has the confidence to speak up assertively. The crew that actively manages workload and calls for help before they’re overwhelmed. These aren’t soft skills. They’re operational necessities—the mechanism that interrupts the chain at step six, before challenge has a chance to weaken. Building that kind of crew competence is a training and culture problem, not a one-off course.

TEM: Detecting Threats Before They Become Accidents

Threat and Error Management (TEM) gives crews a structured way to stay ahead of the chain described above. It’s the practical language competent crews use to articulate what’s happening and what they’re doing about it.

The TEM model is simple but powerful:

1. Identify threats early — weather, fatigue, system anomalies, operational pressure, navigation uncertainty. What could go wrong? Be specific. Name it. Own it.
2. Anticipate the errors those threats are likely to produce — if the weather is deteriorating rapidly, what errors are we most likely to make? Rushing the approach? Missing a callout? Getting behind the aircraft?
3. Detect errors while they are still recoverable — this is where monitoring and cross-checking come in. The monitoring pilot isn’t just there to observe; they’re there to catch errors and bring them to the crew’s attention.
4. Contain the consequences — if an error occurs, act quickly to prevent it from escalating. Stabilise the situation. Buy time if you need it.
5. Recover before the situation becomes irreversible — the competent crew doesn’t just survive; they recognise the problem, address it, and return to a safe state with sufficient margin.

Formally, TEM traces a specific pathway: a threat in the operating environment can lead to a crew error, which can lead to an undesired aircraft state—an unstable approach, a track deviation, an incorrect configuration—which, if not recovered, leads to an outcome. Steps two to five above exist to interrupt that pathway as early as possible, ideally before an error ever produces an undesired aircraft state. This is documented in ICAO’s TEM framework material.

TEM is not theoretical. It is the practical language of competent crews. When a crew can articulate the threats they face and the margins they are protecting, they are far less likely to be surprised by the outcome—and far more likely to catch the chain early, while it is still cheap to break.

How Operators Should Actually Assess Competence

Periodic proficiency checks remain necessary. They’re the foundation of safety. They ensure pilots can meet the technical standards required for their certificate and type rating.

But they are no longer sufficient.

Regulatory and training frameworks such as the FAA’s Advanced Qualification Program (AQP) provide a model for evidence-based and scenario-based assessment that examines the full range of skills contributing to safe decision-making. These assessments look at:

• Decision-making under incomplete information
• Threat recognition and TEM performance
• CRM behaviours in realistic pressure situations
• Manual flying and automation management
• Workload management and prioritisation
• Diversion, abort, and go-around decision-making
• Recognition of personal and crew limitations

The FAA describes AQP as a voluntary, data-informed alternative to traditional training and checking requirements, built around scenario-based training and integrated technical and crew-resource skills. It treats training as continuous rather than episodic—recognising that competence isn’t static; it needs to be maintained, practised, and assessed over time.

Scenario-Based Training vs Tick-Box Checking

The industry must continue shifting the primary evaluation question—from “Did the pilot perform the procedure correctly?” toward a harder, more useful one.

“Did the pilot recognise the threat, understand the evolving situation, evaluate the realistic alternatives, and make the safest available decision—before the last safe opportunity had passed?”

This doesn’t lower technical standards. It raises the standard for judgement. It requires scenario design that includes ambiguity, time pressure, conflicting goals, and imperfect information—the conditions crews actually face.

A few shifts make the biggest difference in practice. Continuous monitoring replaces a single annual snapshot with an ongoing picture built from line data, flight data monitoring, and instructor feedback. Scenario-based training places pilots in realistic, ambiguous situations that require integrated decision-making, not just discrete manoeuvres. Evidence-based training uses data from line operations and safety occurrences to target the specific threats a fleet or route actually faces. And CRM integration means challenge and communication are built into every scenario and every debrief, not taught as a separate module.

None of it works without culture change. Operators need to be honest with themselves about whether it is truly safe to challenge the captain—whether that’s genuinely encouraged, or only tolerated in theory. The answer matters. A safety culture that doesn’t reward challenge is a culture that will eventually pay the price.

What Makes a Truly Competent Pilot?

Every professional pilot eventually faces a moment when continuing feels possible but stopping is safer. Fuel is still adequate. Weather is still legal. The destination is still open. Yet something in the overall picture no longer supports the original plan.

That moment is the purest test of competence.

The pilot who can recognise it, articulate the concern, and act on it—even when the mission feels important and the pressure is real—is the pilot operators and passengers actually need. It takes courage to say, “We’re diverting.” It takes humility to say, “I missed that callout, let’s go around.” It takes professionalism to declare an emergency early.

A truly competent pilot is not defined by any single skill on the list—technical knowledge, handling, CRM, TEM, workload management. It is the integration of all of them, applied consistently, under pressure, without waiting for certainty. The pilot who always presses on is not the pilot you want in the seat when things go wrong. The pilot you want is the one who has the judgement to stop, the courage to challenge, and the wisdom to know that safety isn’t about pride or mission completion. It’s about arriving safely, every time.

Aviation safety investigations repeatedly demonstrate the importance of examining not only what happened, but how operational threats, human factors, decisions and organisational conditions interacted before the outcome. Bodies such as ICAO, through its Annex 13 investigation standards, and national investigation authorities apply exactly this kind of analysis after an occurrence. The patterns described in this article are drawn from that established body of safety literature, not from any single event.

Conclusion: The Best Pilot Is Not the One Who Completes Every Mission

Let’s return to the question we started with: why do experienced, qualified, technically proficient pilots sometimes make decisions that look obviously wrong in hindsight? Not because they lack skill. Because good decision-making is fragile under pressure, and it erodes in small, defensible steps unless a crew is trained—and a culture is built—to interrupt that chain before the last safe opportunity has passed.

Technical proficiency is essential. It’s the bedrock of safe aviation. Without it, you don’t have a pilot; you have someone who’s dangerous. The check ride, the type rating, the recurrent training—these are all absolutely necessary.

But they’re not enough.

At Safe Fly Aviation, we believe aviation safety depends not only on aircraft and procedures, but on disciplined preparation, sound operational judgement and a culture that supports speaking up when the safer decision is to stop.

Passing a check ride proves a pilot can fly the aircraft. Competence proves the pilot can be trusted when the aircraft, the weather, or the plan no longer cooperates.

In modern aviation, that difference is everything.

Frequently Asked Questions

Sources & Further Reading

• ICAO — Aviation Crew Resource Management (ACRM)
• ICAO — Threat and Error Management framework material
• FAA — Advanced Qualification Program (AQP)

Safe Fly Aviation is a global aviation services company providing aircraft charter, helicopter charter, air ambulance, aircraft sales and acquisition, and aviation consultancy. Our work spans aircraft operations, sourcing and aviation support across international markets.

This article is intended for general aviation-information purposes and does not replace applicable regulations, operator SOPs, aircraft manuals, training, or professional guidance.