Bonus Terms Read at 9 pt Shed 18% of Comprehension in 500 Tests

2026-09-28

Across 500 comprehension tests conducted between March and November 2024, reducing the type size of online casino bonus terms from 12 pt to 9 pt lowered median comprehension scores by 18.4% — from 71.2% to 58.1% correct answers on a ten-item assessment covering wagering contribution, maximum bet rules, and expiry windows. The effect was not evenly distributed: participants aged 55 and over lost 26.7% of their baseline comprehension, while the 18–24 cohort lost 9.3%. The finding matters less as a statement about font sizes than as a statement about where the industry has chosen to place the burden of understanding.

What the Test Actually Measured

The study design matters more than the headline number, because bonus-term comprehension is unusually easy to measure badly. Participants were recruited from four English-speaking markets (United Kingdom, Ireland, Canada, Australia) and screened for prior online casino account ownership within the previous 12 months. Each participant saw the same promotional offer — a 100% deposit match up to €200 with a 35x wagering requirement — rendered in one of three type sizes: 12 pt, 10 pt, or 9 pt. The underlying text was identical across conditions, including the 41-word clause specifying that bonus funds carry a 30-day expiry and that deposits via Skrill and Neteller are excluded from the match.

After a two-minute reading window, participants answered ten multiple-choice questions. The questions were deliberately practical rather than pedantic: If you deposit €100 and wager only on slots contributing 80%, how much of the bonus must you turn over before withdrawal? Correct-answer rates fell monotonically with type size, but the drop was steeper between 10 pt and 9 pt (a 9.1 percentage-point decline) than between 12 pt and 10 pt (a 4.0-point decline). That non-linearity suggests a threshold effect rather than a smooth degradation — somewhere around 9 pt, readers stop scanning and start guessing.

Two controls are worth noting. First, participants were given no time pressure beyond the reading window and were not asked to act on the terms, only to understand them. Second, all three conditions used the same line spacing (1.15) and contrast ratio; the study isolated type size, not overall legibility. A separate arm of 120 participants who received the 9 pt text with a bolded wagering clause scored 63.8%, recovering roughly a third of the loss. Formatting interventions, in other words, partially substitute for size.

Why 9 pt Is Not an Arbitrary Benchmark

Nine-point type is not a hypothetical. A 2023 audit of 60 licensed operators in three regulated markets found that 22 displayed bonus terms at an effective size between 8.5 pt and 9.5 pt on desktop, rising to 31 operators on mobile, where responsive layouts frequently shrink legal text below the desktop minimum. The same audit found that median time-on-page for bonus terms was 14 seconds — long enough to scroll, too short to read 1,400 words of conditions.

Regulators have noticed, unevenly. The UK Gambling Commission's 2022 customer-funds and fair-terms guidance does not specify a minimum type size, though it requires terms to be "prominent" and "not misleading." The Malta Gaming Authority's 2019 player-protection directive is similarly silent on point size. Only a handful of jurisdictions — notably several US state regulators following the 2018 Murphy v. NCAA repeal — have written explicit typographic minimums into advertising rules, and those tend to address television disclaimers rather than on-site bonus text. The result is a regime where a term can be fully disclosed and still functionally unread.

The Comprehension Gap Is Not Uniformly Distributed

The 18.4% aggregate figure conceals a distribution that should concern anyone designing for a general audience. Splitting the sample by age band:

18–24 (n=112): 9.3% decline from 12 pt to 9 pt
25–39 (n=168): 15.1% decline
40–54 (n=131): 21.6% decline
55+ (n=89): 26.7% decline

The 55+ cohort also showed the widest variance, with a standard deviation of 19.4 percentage points against 12.8 for the youngest group. That matters for two reasons. First, this cohort holds a disproportionate share of retail and online gambling expenditure in several mature markets — UK Gambling Commission participation data has consistently shown higher average monthly spend among the 45–64 bracket than among 18–24s. Second, presbyopia, the age-related loss of near vision, typically begins around 40 and becomes near-universal by 55; a 9 pt clause read at arm's length on a phone is precisely the scenario optometry research identifies as the hardest common reading task.

There is a defensible counterargument: younger players, who lost least from the size reduction, are also the cohort most likely to read terms on a phone at default zoom, where 9 pt is closer to 12 pt in effective angular size. If the study's 9 pt condition was rendered on a desktop monitor at typical viewing distance, it may overstate the mobile penalty for young users and understate it for older ones. The authors acknowledge this; a follow-up with device-controlled conditions was scheduled for early 2025.

The Wagering-Contribution Question

One item in the assessment deserves separate attention because it produced the largest gap between conditions. Participants were asked which of four game categories contributes 100% toward wagering. At 12 pt, 68.4% answered correctly. At 9 pt, 41.2% did. The correct answer — slots, with most table games contributing between 10% and 20% and live dealer often excluded entirely — is exactly the kind of term that determines whether a player ever clears the bonus, and it is exactly the kind of term most often buried in a sub-clause.

This is not a comprehension failure in the abstract. A player who misreads contribution rates is not merely uninformed; they are exposed to a materially different expected value than they believe they are. If a €200 bonus with a 35x requirement is wagered entirely on a game contributing 10%, the effective requirement is 350x, not 35x. The gap between those two numbers is the difference between a bonus that is theoretically clearable and one that is, for practical purposes, not.

What the Numbers Imply for Operators and Regulators

The study's authors are careful not to overclaim. Comprehension of a ten-item quiz is not the same as informed consent, and a player who scores 58% may still be adequately warned if the missed items are peripheral. But the wagering-contribution result is not peripheral. Nor is the expiry window: at 9 pt, only 47.9% of participants correctly identified the 30-day deadline, against 74.1% at 12 pt.

Three implications follow, none of them comfortable. First, if regulators accept that disclosure must be effective rather than merely present, then type size becomes a compliance variable, not a design preference — and the current patchwork of "prominence" language is inadequate to enforce it. Second, operators who argue that terms are "fully disclosed" are making a claim about availability, not about comprehension, and the two diverge measurably at 9 pt. Third, the formatting arm of the study suggests a cheaper fix than larger type: bolding the wagering clause recovered 5.7 percentage points, roughly a third of the total loss, at no cost to page layout.

The open question is not whether 9 pt terms are harder to read — that is now measured — but whether the industry's disclosure model can survive the finding that its most consequential clauses are the least reliably understood. If a term is legible only to a reader who already knows to look for it, the disclosure has shifted from informing the player to protecting the operator. The 500 tests do not resolve that question. They make it harder to avoid.