Evidence vs. Rhetoric: Why the Jury Is Still Out on Gambling Harm Data

By Simon Barff, Managing Director, CLMS

“I think we start with the evidence – because the evidence doesn’t match the rhetoric.”

The recent release of the Gambling Survey for Great Britain (GSGB) has once again raised questions about the way gambling harm is measured in this country. According to the GSGB, problem gambling rates have risen. Yet, when you dig into the detail, as Andrew Tottenham of Tottenham & Co. did in his September Tottenham Report, it becomes clear that the numbers may be more about methodology than reality.

Confirmation Bias in Action

Tottenham compares the Commission’s interpretation of the GSGB to a TV experiment called The Jury. Jurors, unfamiliar with the evidence, were quick to make up their minds and then twisted the facts to fit their conclusions – a textbook case of confirmation bias.

Tottenham argues that the Gambling Commission did something similar: “The Commission seems eager to treat [the experiments] as proof that the GSGB is the superior instrument… Meanwhile, the strongest and clearest finding, the topic salience effect, was largely downplayed.”

In other words, the conclusion came first, and the evidence was arranged to support it.

What the Experiments Really Show

The Sturgis experiments underpinning the GSGB review tell a more complicated story:

Topic salience bias: When a survey is labelled “about gambling,” it attracts more gamblers and more engaged gamblers – inflating harm rates.
Interview method bias: Online self-completion surveys may elicit more admissions, but comparisons to the HSE (which used face-to-face with self-completion) don’t hold up.
Activity list length: Hardly any impact.
As Tottenham concludes: “The experiments shed some light, but not enough to deliver a final verdict… The jury is still out.”

What the Data Really Shows in Practice

At CLMS, we have the advantage of working with hard evidence from over 35,000 connected Category C machines across the UK – one of the largest aggregated datasets in the industry. That data paints a consistent picture:

Stakes have remained stable over the last 15 years, once inflation is accounted for.
Value of play, time on machine, and cash collected show no significant spikes or shifts.
Meanwhile, game design has improved markedly, with Return to Player percentages rising from the mid-70s to the high-80s. In other words, players are enjoying a better value entertainment experience than ever before.
This is not an industry spiralling into harm – it’s an industry that has remained remarkably consistent, even as technology has evolved.

The Risk of Mismatched Narratives

If policy is based on skewed or inflated data, we risk making the wrong interventions. As Tottenham warns:

“When limitations are glossed over and uncertainties minimised, we risk building policy on shaky ground.”

At CLMS, we believe the UK’s land-based small stake, small prize sector remains one of the safest gambling environments in the world. The data backs that up – not just rhetoric.

Moving Forward

The stakes are high – not just for operators, but for policymakers, regulators, and the communities they serve. That’s why we’ll continue to use our data to bring transparency to the debate and ensure that decisions are made on the basis of evidence, not assumptions.

Because when it comes to gambling policy, the jury really is still out – and the facts need to speak louder than the narrative.

Read Andrew Tottenham's excellent article here

Return to the previous page

2 October 2025

See all news...