The Language of Safety: Risk Scores – When Numbers Mean Lives.

The safety manager responded that it was always extreme.

Extreme meant fatality almost certain, so I asked if they had experienced a fatality with almost every work at heights task.

“No, thank goodness. No incidents at all actually.”

“Then why is it scored extreme?”

“Because falls can kill people.”

They were right, but this response didn’t take into account the different work at heights scenarios, locations and existing controls.

This exchange reveals one of the most common and dangerous failures in risk assessment: scoring based on worst-case scenarios rather than realistic probabilities.

When Numbers Lie About Danger

Risk scores aren’t just numbers on a matrix. They’re the language that tells your organisation where to direct its protective resources.

When those scores are inflated or deflated, they don’t just distort your documentation – they actively misdirect your protection efforts.

AS/NZS ISO 31000 defines risk score as “the magnitude of a risk expressed in terms of the combination of consequences and their likelihood.” It sounds straightforward until you watch how it is frequently applied.

The Art and Science of Honest Scoring

Risk scoring requires both rigorous methodology and clear-eyed judgment. Get either wrong and your entire risk management system becomes compromised.

Likelihood: Probability, Not Possibility.

I once watched a risk assessment where every hazard was scored as “likely” or “almost certain”, regardless of historical evidence.

The justification? “It could happen anytime.”

This approach isn’t just wrong – it’s dangerous. When everything is high priority, nothing is.

To assess likelihood accurately:

1. Examine Actual Exposure – How many people interact with this hazard, how often and for how long?
2. Review Historical Data – Whats the actual incident rate in your organisation and similar operations?
3. Consider Near Misses – These are powerful predictors of future incidents.
4. Account for Existing Controls – Are you assessing raw risk or risk with current controls in place?

A manufacturing plant I worked with had rated “caught in conveyor” as “almost certain” despite having no incidents in 15 years of operation. This inflated likelihood pushed up the overall risk score, diverting resources from managing more pressing risks, like chemical exposures that were causing regular minor injuries and consequently had the potential to lead to long term health effects in their workers.

Consequence: Typical, Not Theoretical

The most common risk scoring error? Rating every consequence based on worst-case scenarios.

“Well, someone could die if…”

This thinking leads to risk matrices filled with extreme consequences for even minor hazards. For example, whilst technically someone could strike their head and possibly die from tripping on uneven ground, the most likely consequence would be either no or minor injury. The result from recording the most severe consequence? Critical resources directed to unlikely scenarios while activities involving more serious consequences may continue unabated.

To assess consequences accurately:

1. Focus on Typical Outcomes – What happens most often when this event occurs?
2. Consider Multiple Consequence – Types Don’t just assess injury – consider operational, financial and reputational impacts.
3. Use Industry Experience – What do similar events typically cause in comparable operations?
4. Apply Reasonable Worst-Case – Not the theoretical worst case, but the reasonably worst outcome.

A construction company I consulted for had scored a paper cut as having “major” consequences because “it could get infected and lead to sepsis.” This is not the most likely outcome. Potentially true, but reasonably unlikely.

The Matrix: Tool or Trap?

Risk matrices often become risk management theatre – impressive documents that create the illusion of protection without delivering it.

Effective matrices:

1. Use Clear Definitions – Every likelihood and consequence level must have unambiguous meanings.
2. Reflect Business Reality – A matrix for a large mining operation will differ from one for a small office business.
3. Drive Proportional Response – High scores must trigger meaningful actions, not just documentation.
4. Balance Detail and Usability – Too simple and they miss nuance. Too complex and they’re ignored.

I recently reviewed a risk management system with a beautifully detailed 7×7 matrix that no one actually used. The shop floor relied on an unofficial 3-level system: “no worries,” “be careful,” and “stop work now.” It may seem simplistic, but for the operations and importantly the people conducting them, this was an effective tool that more accurately reflected their level of risk and how to adequately control it.

The Consequences of Inaccurate Scoring

When businesses get risk scoring wrong, several critical failures follow:

1. Resource Misallocation – Time money and attention directed to the wrong risks.
2. Risk Assessment Fatigue – When everything is extreme people stop taking assessments seriously.
3. False Security – Low scores for genuine high risks leave workers unprotected.
4. Decision Paralysis – Too many high-scoring risks overwhelm management response capacity.

I was working at a site that had so many “extreme” risks in their register that management were forced to arbitrarily rank them in order to gain the CAPEX, time and resources to address them in some sort of order. This is an understandable response, but the outcome was that a potential multiple fatality risk slipped through and was ranked significantly lower than other items. Fortunately, when this equipment failed, it was fortuitously caught (not long) before it created a potentially large explosion. Too many extreme risks resulted in a more dangerous work environment.

Making Numbers Tell the Truth

Risk scores should tell a story about where your real dangers lie. To ensure yours do:

1. Calibrate Your Team – Train assessors to score consistently across departments.
2. Test Against Reality – Compare scores against actual incident history to verify accuracy.
3. Review Outliers – Exceptionally high or low scores deserve special scrutiny.
4. Revisit Regularly – Risk scores change as operations, training and controls evolve.

Risk scoring isn’t about creating impressive documentation. It’s about directing protective resources where they’ll save the most lives and prevent the most harm.

How we write that’s different

Write Safe AU specialises in creating risk scoring systems that reflect reality rather than fear.

Our approach helps businesses develop scoring methodologies that balance statistical accuracy with practical application. We calibrate matrices to your specific operations, ensuring that the numbers mean something real to your people and drive appropriate protective actions.

This scoring-focused methodology means your risk assessments don’t just satisfy auditors, they genuinely protect your people by directing resources to the dangers that pose the greatest actual threat. Because when your numbers tell the truth about risk, your protection efforts focus where they matter most.

Because honest risk scores don’t just organise priorities, they save lives.

This is The Language of Safety. Clear documentation. Safer workforce.

Write Safe AU