A £50 cash-out PIN at 10pm cut split payments by 33% in the first eight weeks of a controlled trial across three UK-licensed operators, according to figures shared with us ahead of publication. The mechanism is deliberately dull: a single £50 withdrawal needs a PIN confirmation after 22:00, and that small piece of friction appears to be enough to stop players fragmenting one withdrawal into three or four smaller ones. Below is what the data actually shows, where the effect is strongest, and why the operational case is more interesting than the compliance one.

What the trial measured

The trial ran from 6 January to 2 March 2025 across three operators with a combined 410,000 monthly active depositors. Players were split so that roughly half sat in a treatment group where any cash-out of £50 or more requested between 22:00 and 06:00 required a six-digit PIN re-entry. The control group saw the existing flow: request, confirm, done.

Split payments were defined as two or more withdrawal requests from the same account within a 30-minute window, totalling less than £200. That definition matters. It excludes the legitimate pattern of a player taking £300 out in two chunks because of a per-transaction cap, and it excludes the player who requests £20 on Monday and £20 on Friday.

The headline: 33% fewer split-payment events in the treatment group, adjusted for baseline differences in account age, deposit history and prior withdrawal behaviour. Absolute numbers were 11.4 split events per 1,000 withdrawal sessions in control, against 7.6 per 1,000 in treatment. The effect held at 31% when the analysis was restricted to accounts with a verified affordability check on file, so it is not an artefact of a few large accounts skewing the mean.

Why 10pm and why £50

The 22:00 threshold was chosen because the operators' own data showed withdrawal requests clustering between 22:00 and 00:30, with a secondary bump after 01:00. The £50 floor is a compromise: low enough to catch the fragmentation pattern, high enough that the friction does not land on the £10–£20 withdrawals that make up the bulk of routine cash-outs. At £30 the pilot produced almost no measurable effect and a noticeable dip in same-session completion; at £100 the effect was larger but the sample thinned out too much to be useful.

Six digits rather than a password was deliberate. A password invites reuse and lockouts; a PIN the player has already set for login is a familiar action that takes about four seconds. Median time from request to confirmed withdrawal in the treatment group rose by 11 seconds. Median time to abandonment — the player closing the cashier without confirming — rose from 2.1% to 3.4% of sessions, which is the number that tells you friction is doing work rather than just annoying people.

The mechanism is not the one compliance teams assume

The intuitive read is that the PIN interrupts a player in a heightened state and gives them a moment to reconsider. That is part of it, but the operator data points somewhere less flattering: the PIN mainly disrupts automation.

A meaningful share of split payments come from players running a routine — a saved withdrawal amount, a habitual sequence of taps, often on mobile with biometrics already primed. The PIN breaks the sequence. In session recordings reviewed by two of the operators, the most common post-PIN behaviour was not abandonment but consolidation: the player returned to the cashier and requested the full balance in one go.

That has a direct cost implication. Split payments are expensive to process. Card and open banking withdrawals carry per-transaction fees that do not scale down with amount, and each additional request adds to the manual review queue, particularly for accounts flagged for enhanced checks. One of the three operators put the saving at roughly £0.14 per avoided split event across payment costs and review time — small per player, but at 410,000 monthly actives it is a five-figure monthly line item, not a rounding error.

Where the effect disappeared

The 33% is an average, and averages hide the useful detail. Two segments showed no statistically significant change:

• Players with a withdrawal limit already set below £50. They were excluded from the treatment by design, but the control group's behaviour in this segment was flat anyway, suggesting self-imposed limits already suppress fragmentation.
• Desktop sessions. The effect was concentrated almost entirely on mobile, where the cashier flow is shorter and the saved-amount habit is stronger. Desktop split payments fell by 6%, within the confidence interval.

There was also a novelty decay worth flagging. Week one showed a 41% reduction; by week eight it was 27%. The operators involved are watching whether that settles or continues to fall. If it decays toward zero, the intervention becomes a one-off nudge rather than a standing control — which changes the case for keeping it permanently.

The responsible gambling question nobody has settled

Here is the uncomfortable part. Fewer split payments is a good operational outcome. It is not obviously a good player-protection outcome, and the two should not be conflated in the write-up.

The trial's own secondary measure — the share of treated players who subsequently set a deposit limit or took a timeout — showed no significant movement. That is a neutral result, not a positive one. A friction that reduces payment costs without changing help-seeking behaviour is a payments optimisation wearing a safer-gambling badge, and operators should be honest about which one they are buying.

The counter-argument is that fragmentation itself is a marker. Players who split withdrawals repeatedly are more likely to be doing so to keep individual amounts below a threshold that would trigger checks, or to manage a bank balance in real time in a way that suggests distress. If that is true, then reducing fragmentation reduces a signal the operator relies on. Neither operator we spoke to could say whether split-payment flags feed their safer-gambling models; one admitted the interaction had not been tested.

There is a straightforward fix, and it is not being deployed. Instead of suppressing the split, log it. A PIN confirmation that also records the reason for the second request — a free-text field, or a two-option prompt — would preserve the behavioural signal while keeping the payment benefit. That is a design question, not a technical one, and it is the obvious next trial.

What to watch

The 33% figure will be quoted in vendor decks within a quarter, stripped of the mobile-only concentration, the week-eight decay and the neutral player-protection result. The more useful question is whether the effect survives a longer window and whether it holds for players who are already in a withdrawal-limit regime — a group the trial largely excluded and the one most worth understanding.

If you are running a cashier flow and considering this, the £50/22:00 configuration is a reasonable starting point, but budget for the decay and instrument the abandonment rate from day one. A friction that annoys without redirecting is just a slower cashier. And if the PIN prompt is going to exist anyway, the field that asks the player why they are splitting costs almost nothing to add — the fact that nobody in this trial added it is the most telling number in the dataset.