A player who sets a $20 deposit limit mid-session — or has one imposed by an operator's safer-gambling trigger — is more likely to route money elsewhere than to stop. New behavioural data puts the share of players who top up through a second account, an e-wallet, or a different brand within 24 hours of a limit taking effect at 44%. That figure is the uncomfortable core of the deposit-limit debate: the tool works for the players who accept it, and leaks badly for the ones who don't.

The data, drawn from a sample of roughly 12,000 accounts across three licensed markets over an 18-month window ending in March, tracked what happened after a limit of $50 or less was applied. Of the 44% who topped up elsewhere, most did so within four hours. The average top-up was $180 — nine times the limit that had just been set.

Why a low limit is the leaky one

There's a counterintuitive pattern in the numbers. Players with limits above $200 stuck with them about 78% of the time. Under $50, that retention fell to 56%. The tighter the constraint, the stronger the incentive to work around it.

Part of this is mechanical. A $20 limit made sense against a $500 monthly budget. It makes no sense against a player who was depositing $400 a week, and it's those players who generate most of the workarounds. A limit that doesn't match the player's actual spending pattern reads as an obstacle rather than a guardrail — and obstacles get routed around.

The second-account problem

Operators can detect a duplicate account at the same brand. They're much weaker at spotting a player opening an account at a competitor, or moving to a crypto wallet, or asking a friend to deposit. Cross-operator data sharing exists in a few jurisdictions — the UK's GamStop and Australia's BetStop are the best-known examples — but they cover self-exclusion, not deposit limits. A $20 limit at Brand A says nothing to Brand B.

What the 44% actually tells us

It would be easy to read this as an argument against deposit limits. That's the wrong conclusion, and the data doesn't support it. Among players who did stay within their limit, average monthly spend fell by 31% over the following six months, and session length dropped by about a fifth. For that group, the tool did exactly what it's designed to do.

The problem is selection. Limits are most effective for players who set them voluntarily and least effective for players who have one imposed after a risk flag. The 44% figure is concentrated heavily in that second group — the players you'd most want the limit to protect.

Where the fix probably sits

Three levers show up repeatedly in the research:

• Limit-setting at onboarding, before the first deposit. Players who pick a limit cold tend to set it closer to their real budget than players who set one mid-session after a loss.
• Cooling-off periods on limit increases. A 24-hour delay before a raised limit takes effect is standard at most major operators now; a delay before removing a limit entirely is less common.
• Cross-operator visibility. The technical infrastructure exists. The commercial and legal appetite doesn't, outside a handful of markets.

None of these solve the e-wallet and crypto route, which is where a meaningful slice of the 44% ends up.

The question nobody wants to answer

If a $20 limit pushes four in ten players toward an unregulated or less-regulated channel, is the limit protecting them or just moving the risk somewhere the operator can't see it? The honest answer is that we don't know yet — the 44% figure measures where the money went, not what happened to the player afterward. That's the study someone should run next. Until then, operators rolling out aggressive low-limit defaults are making a bet on player behaviour that the evidence doesn't fully back.