Skill scores predict 2 in 3 hiring outcomes, chance fills the rest

Recruiters in Poland's engineering and construction sectors have spent the past decade layering competency frameworks, structured interviews, and skills assessments onto the hiring process. Yet even the most disciplined talent teams still watch a meaningful share of hires succeed or fail in ways their pre-hire scores never predicted. If skill signals explain roughly two in three outcomes, what accounts for the remaining third — and can anything be done about it?

The honest answer sits at the intersection of behavioural psychology and the messy reality of infrastructure work: hiring is a decision made under uncertainty, and the parts of a job that matter most are often the parts least visible at the point of selection.

The 64% That Scores Can See

The claim that skill scores predict about two-thirds of hiring outcomes is not a marketing line; it is roughly what industrial-organisational research has found for decades. Frank Schmidt and John Hunter's landmark 1998 meta-analysis of 85 years of personnel selection research put the validity of general mental ability tests at around 0.51 for job performance, with structured interviews and work-sample tests landing in similar territory. Stack a few valid predictors together and you approach the 0.6–0.7 range — a genuine edge over chance, and a long way from certainty.

In Polish construction and transport, this maps onto familiar tools: technical exams for site engineers, licence and certification checks for heavy equipment operators, design-portfolio reviews for infrastructure planners. These instruments do real work. A candidate who cannot read a geotechnical survey or size a drainage culvert will struggle regardless of how well they interview. But the residual third is where projects quietly go wrong.

Why the Residual Third Resists Measurement

The unmeasured portion is not random noise. It is systematic, and it clusters around a handful of behavioural traits that are extraordinarily hard to assess in a two-hour interview.

Risk appetite under genuine uncertainty

Kahneman and Tversky's work on loss aversion established that people weigh losses roughly twice as heavily as equivalent gains — and that this asymmetry distorts decisions in predictable ways. On a construction site, that distortion shows up as the supervisor who refuses to halt a pour despite a marginal weather forecast, because stopping feels like a visible loss while continuing feels like a normal day. A skills test can confirm someone knows the safety protocol. It cannot tell you whether they will invoke it at 4 a.m. with a deadline looming and a regional manager on the phone.

The variable-ratio problem in project work

Behavioural psychology's most durable finding about persistence is the variable-ratio schedule: behaviour reinforced unpredictably is far more resistant to extinction than behaviour reinforced consistently. B.F. Skinner's work on intermittent reinforcement explains why some engineers thrive on the lottery-like rhythm of infrastructure delivery — months of routine followed by a crisis that consumes a weekend — while others burn out. Hiring scores rarely capture this. A candidate's tolerance for unpredictability is closer to temperament than to competence, and it predicts retention and decision quality under pressure.

Competitive play and team dynamics

Transport and construction are team sports played with contracts, not balls. Research on competitive environments — from tournament theory to studies of how rivalry affects risk-taking — suggests that people behave differently when they perceive themselves to be in a contest. On a bid team, that can produce sharper estimates or reckless lowballing. In a site office, it can produce either healthy urgency or information hoarding. Structured interviews probe for "teamwork" in the abstract; they seldom recreate the conditions under which competitive instincts actually surface.

What a Study of Site Failures Suggests

One instructive reference point comes from the body of research on engineering project overruns. Bent Flyvbjerg's large-sample studies of megaprojects — including transport infrastructure — found that cost overruns are the norm rather than the exception, and that they are driven less by technical miscalculation than by optimism bias and strategic misrepresentation in the planning and approval stages. Those are behavioural failures, not skills failures. The people who produced them were, by any competency framework, qualified.

The implication for hiring is uncomfortable but useful: the traits that generate overruns — overconfidence in forecasts, reluctance to escalate bad news, anchoring on an approved budget — are the same traits that interviewers often read as confidence and decisiveness. A skills score will happily endorse the candidate who most needs a sceptical colleague beside them.

Designing for the Third You Cannot Score

Forward-looking firms in Poland's infrastructure and logistics sectors are starting to treat the residual third as a design problem rather than a selection problem. Three moves recur.

Probe decisions, not traits. Rather than asking whether a candidate is "calibrated" or "risk-aware," present a concrete scenario — a subcontractor's work fails inspection two days before a milestone — and ask what they would do, then ask what would change their mind. The follow-up question is where loss aversion and escalation reluctance reveal themselves.

Use structured work samples with embedded ambiguity. A design exercise with a deliberately incomplete brief tests not only technical skill but tolerance for uncertainty. How a candidate responds to missing information — asking, assuming, or freezing — is a behavioural signal that no exam captures.

Build the missing third into the team, not the hire. If scores cannot reliably identify the person who will halt a risky pour, pair the decisive site lead with a colleague whose role is explicitly to challenge forecasts. This is the practical version of what project-management research calls reference-class forecasting: institutionalise the scepticism rather than hoping to recruit it.

For Poland's pipeline of road, rail, and logistics investment, the stakes are concrete. The two-thirds that scores predict will keep improving as assessment tools sharpen. The remaining third will not yield to better tests, because it is not a knowledge deficit — it is a behavioural one, shaped by how people weigh losses, respond to unpredictability, and compete. The firms that get ahead will be the ones that stop treating that third as a hiring failure and start treating it as an operational design brief.