Why competent people still make the wrong choice
How pressure, context and familiarity undermine good judgement in capable teams.
Let me start with something you already know, even if we rarely say it out loud.
Most serious incidents are not caused by people who are careless, lazy, or untrained.
They are caused by people who are competent, experienced, and trying to do the right thing.
That truth makes incidents harder to talk about. It’s more comfortable to blame ignorance than to accept something more unsettling: that knowing what to do is not the same as doing it when it matters.
I’ve stood on enough sites, spoken to enough supervisors, and read enough investigation reports to see the same pattern repeat. The paperwork is sound. The records are up to date. The procedure is clear. And yet, at the critical moment, the wrong choice is made.
Under stress, we don’t rise to the level of our training. We fall to the level of our practice.
Not because the person didn’t know better.
Because the situation did not resemble the training that prepared them.
Knowledge collapses under pressure
Most training is delivered in what I’d call a cold context.
Quiet room. Predictable pace. No competing demands. No real consequence.
Real work is different.
Real work is noisy, interrupted, time-pressured, and socially complex. People are thinking about production targets, colleagues watching, radios crackling, weather changing, and whether stopping the job will cause an argument or a delay.
Under those conditions, the brain does not calmly retrieve information from a training slide. It simplifies. It shortcuts. It falls back on habit and pattern.
These shortcuts are rarely reckless. They are usually shaped by production pressure, time constraints, and a desire to keep work moving. In the moment, they often feel sensible - even responsible. The problem is not that people take shortcuts. It’s that they take them without having practised where those shortcuts fail.
That’s not a flaw. It’s how human beings are wired.
Under stress, we don’t rise to the level of our training. We fall to the level of our practice.
This is why post-incident interviews so often include phrases like:
- “It didn’t feel that serious at the time.”
- “I thought it would be fine.”
- “I’ve done it this way before.”
These aren’t excuses. They are clues.
The confidence trap
Experience, oddly, can make this worse.
Competent people build mental models: We can say “I’ve seen this before”.
Those models are usually helpful - until a situation is just different enough to matter.
Hazards don’t announce themselves clearly. They emerge gradually, or they look familiar when they are not. A valve sounds slightly different. A space feels tighter than expected. A routine task happens at the end of a long shift.
The risk is not ignorance.
The risk is misplaced confidence.
Traditional training rarely challenges this. It teaches the correct answer, but not the moment where a wrong answer feels reasonable.
Why refresher training doesn’t fix it
When incidents repeat, the default response is often “more training”.
The intention is good. The outcome is usually disappointing.
Why? Because repetition of the same format does not change behaviour. It reinforces familiarity, not judgement. People sit through it, nod along, sign the sheet - and return to work unchanged.
What’s missing is not information.
What’s missing is experience under pressure.
People need to feel the moment where a choice becomes ambiguous. Where stopping feels costly. Where continuing feels normal. Where the consequences are delayed just enough to be tempting.
If training never puts people in that moment, we shouldn’t be surprised when the real world does - and they’re unprepared.
Choice, not compliance
This is the part many organisations struggle with.
Safety is often framed as compliance: follow the rule, complete the task, tick the box. But incidents don’t happen at the point of compliance. They happen at the point of choice.
Those choices are rarely binary. They are subtle, contextual, and human.
And they are precisely the choices most training avoids, because they are uncomfortable to simulate and difficult to measure.
Yet if we are serious about preventing harm, these are the moments we must train for.
A different starting point
If competent people still make the wrong choice, the question is not “how do we explain it better?”
The question is:
How do we let people practise the right choice before the real world demands it?
That requires training that feels close enough to reality to matter. Not perfect realism. Not expensive theatre. Just enough pressure, consequence, and ambiguity to trigger the same mental processes that operate on the job.
That’s where real improvement begins.
In the next chapter, we’ll look at what effective practice actually looks like - and why realism is less about visuals, and more about decision-making under constraint.