The Bell Curve Lie: Why Forced Rankings Break the Teams They’re Meant to Measure

Jul 2

5 min read

I’ve sat in enough calibration sessions to know what they really are.

A manager presents their ratings. The ratings go up the chain and someone from HR looks at the distribution and the numbers are nudged. Not because the evidence changed or because a new observation came to light. The shape has to fit the curve so the scores get adjusted to fit the shape.

That’s not performance management. That’s utter nonsense and everyone knows it.

I’ll give you a specific example.

At one organisation, I had a target to make two people redundant. They had scored lowest against the HR criteria. The decision was justified, documented, and done. Less than a month later, the annual performance review opened.

I submitted my scores. No low scorers. HR came back and told me I needed to moderate down.

I refused.

It was escalated upwards, formally and with urgency. My rationale was simple. I had just made the low scorers redundant. They no longer worked there and I wasn’t going to manufacture underperformers to satisfy a distribution model when the actual underperformers had already left the building.

HR, reluctantly, left my scores where they were.

That moment has stayed with me. Not because I won the argument, but because the argument had to happen at all.

Teams Aren’t Populations

The bell curve, in theory, makes sense. If you have enough people, performance should distribute naturally. A few exceptional. A few struggling. Most somewhere in between. It’s how statisticians think about populations. The problem is, you’re not managing a population. You’re managing a team. Teams don’t behave like populations.

A team built well, recruited carefully, led consistently, and developed over time can shift the whole curve to the right. Standards should rise and expectations will normalise. What looked like exceptional two years ago becomes baseline today. That’s not a distribution problem. That’s what good leadership looks like.

Forcing those ratings back into a bell shape punishes that progress.

What Moderation Actually Does

Let’s be honest about what moderation actually does.

The high and middle performers take the hit. The high achievers scored a nine but the curve needs fewer nines, so they become a seven. Most middle scorers stay where they are but some will get nudged down into low territory. Those people go home knowing they delivered, knowing their manager recognised it, and knowing the organisation capped it anyway.

Do you think it doesn’t affect what they bring next year?

The low performers, meanwhile, often land exactly where they always land. The moderation doesn’t move them. The system rarely has the courage, to deal with persistent underperformance directly. So the curve is met at one end and silently tolerated at the other.

They’ve all seen this before. They’ll see it again. They roll their eyes, sign the form, and carry on. The performance review becomes background noise. Something to get through rather than something to use.

The real question isn’t whether the bell curve fits your data. The real question is whether your performance framework is changing behaviour, developing people, or making decisions that matter. If moderation is overriding the evidence your managers are generating, then the answer to all three is probably no.

The System Trains Managers to Stop Telling the Truth

There’s a deeper problem too.

When a manager knows their ratings will be moderated, they start calibrating differently. They stop scoring what they see and start scoring what they think will survive the process. The honest, careful, evidence-based judgment they could bring gets replaced by a kind of strategic positioning.

Over time, the system teaches managers to ignore their own honest assessments. It trains them, quietly, consistently, to game the process rather than engage with it.

That’s the opposite of what you want. Leadership requires judgment. Performance management should sharpen that judgment. Forced distribution undermines judgement and leadership.

Amy Edmondson’s work on psychological safety gives us a useful lens here. Teams perform best when people feel their honest contribution, whether as leaders or individuals, is seen and valued (Edmondson, 2018). A moderated rating system tells your high performers the opposite. It tells them that the organisation’s need to hit a shape matters more than what they actually did.

That’s a message, and when it lands don’t be surprised when the best people start listening to it.

What Good Looks Like Instead

So what should happen instead?

Performance assessment should be evidence based, criteria led, and consistently applied. If someone is performing at the highest level, that rating should stand regardless of how many others are performing at the same level. A high performing team that produces five exceptional individuals shouldn’t be forced to invent two average ones to balance the model.

Criteria matter here. If the standards are clear, if everyone knows what exceptional looks like, what meeting expectations looks like, and what falling short looks like, then the ratings reflect reality rather than politics. You’re not comparing people to each other. You’re comparing them to a standard.

That’s a fundamentally different exercise and it’s far more honest.

Honest Assessment Is Harder, Not Softer

None of this means performance assessment should be soft. The opposite.

Honest, evidence based scoring is more demanding than fitting a curve. It requires managers to make real judgments and justify them. It requires organisations to act on what those judgments reveal, including the difficult conversations about persistent underperformance that moderation quietly allows everyone to avoid.

The bell curve gives the impression of differentiation without requiring anyone to be genuinely accountable for the conclusions. You haven’t called someone a poor performer. The distribution did. For the person on the receiving end of an unjustified low score, that distinction means nothing. They know what they delivered. So do their colleagues.

So ask yourself what someone with an unjustified low score is likely to do next.

They leave. The best people have options. If the organisation has just told them, without words that their contribution doesn’t matter, they’ll find somewhere that values it. This is the most damaging outcome, and it’s silent. No exit interview captures “I was moderated down unfairly.” It just looks like natural attrition.

They disengage. If leaving isn’t immediately viable, they recalibrate their effort to match the score they were given. Why deliver a nine if you’ll be rated a six? The system has just taught them that output and reward are decoupled.

They grieve it formally. Some will raise a grievance or lodge an appeal. That costs the organisation time, management energy, and potentially legal exposure, all to defend a score that wasn’t based on evidence in the first place.

They talk. Quietly, persistently, to colleagues. Nothing erodes team culture faster than a credible person saying “I was marked down and nobody could tell me why.” It spreads.

None of these outcomes are accidental or unforeseeable. The system produced them, and they’re predictable. That’s not leadership. That’s a spreadsheet making decisions that people should be making.

When the System Consumes Itself

Stephen Covey wrote about the danger of subordinating principle to procedure, of following the system so faithfully that you forget what the system was supposed to achieve (Covey, 1989).

Performance management exists to develop people, recognise contribution, and address underperformance. When moderation undermines all three of those purposes, the system has failed.

The question isn’t whether the curve should be met. The question is whether your organisation has the clarity and the courage to score honestly, act on what it finds, and trust its managers to tell the truth about the people they lead.

If the answer is no, the bell curve isn’t your problem. The culture is.

“It’s what you do. Not what you’re called.”

Published by The Everyday Leader

Written by Craig Bromage

www.theeverydayleader.online