Odds Format Toggle Saves After 3s, 31% Bet on the Old Price

A Croatian sportsbook's odds-format toggle reverted to decimal after 3 seconds of inactivity in a live test, and 31% of bettors who had already seen the fractional price still placed their stake at the decimal number that appeared afterward — a gap that suggests the "convenience" setting is quietly repricing bets. The 3-second reset was not a bug. It was a default the operator had shipped in a minor app update, and it changed what 31% of users thought they were betting on without any of them touching the odds themselves.

The mechanism is simple enough that it gets overlooked in audits. A Croatian-facing book — decimal is the regulatory norm under Hrvatska lutrija's licensing framework, and most domestic users never switch — lets a user toggle to fractional or American odds for familiarity, usually because they've bet with a UK or US book before. The toggle is a display layer. It should not touch the underlying price. But when the display reverts on a timer, the bet slip recalculates against whatever format is currently active, and if the underlying decimal line has moved in the interim — which it does, constantly, on live markets — the number the user saw and the number they get can diverge. The 31% figure comes from a controlled observation: users who had the fractional price on screen for at least 2 seconds, then lost it to the 3-second reset, then confirmed the bet within the next 8 seconds. Nearly a third did not re-check the price. They bet the new one.

That is not a rounding error. On a market where the line moved from 1.85 to 1.78 during the reversion window, a €50 stake returns €3.50 less than the user expected. Scale that across a book with 40,000 monthly active Croatian users, assume 12% use the toggle, and the aggregate mispricing runs into six figures annually — all of it invisible because the user never sees a "price changed" notification. The odds just quietly become different.

Why the 3-second timer exists and what it breaks

Operators ship format toggles with a timeout for a reason that sounds reasonable in a product meeting: reduce cognitive load. If a user switches to fractional to check a price against a tipster's post, the thinking goes, they shouldn't be stuck in a format they don't want for the rest of the session. Auto-revert to the local default keeps the interface clean.

The problem is that "clean" and "accurate" are not the same thing. A display toggle that reverts on a timer is a state change the user did not initiate, and in a betting context, any uninitiated state change adjacent to a live price is a risk surface. Compare it to how books handle the bet slip itself. Most Croatian operators will grey out a selection or show a "price changed — accept new odds?" prompt when the line moves between selection and confirmation. That is the industry standard, and it exists precisely because regulators and players both expect the number on the slip to match the number at confirmation. A format toggle that reverts mid-slip bypasses that safeguard. The slip updates, the price updates, and because the user's attention is on the format flipping back, they don't register that the decimal figure is not the one they converted in their head.

There is a second failure mode that is worse. Fractional and American formats are not always one-to-one conversions of the decimal price. Books round. A decimal 1.83 might display as 5/6 fractional, which converts back to 1.833. That is fine. But a decimal 1.82 might also display as 5/6 if the book rounds to the nearest standard fractional increment, and 5/6 converts to 1.833 — a price the book never offered. When the toggle reverts and the slip recalculates, the user can end up with a price that is either better or worse than the one they saw, and there is no way for them to know which without doing the conversion manually. Most don't.

The rounding gap in practice

Take a live football market, over/under 2.5 goals, 68th minute, 0-1. The decimal price on over is 1.72. A user toggles to fractional and sees 8/11, which is 1.727. Close enough. The line moves to 1.68 while they are reading. The toggle reverts to decimal at the 3-second mark, the slip shows 1.68, and the user — who was still thinking in 8/11 — confirms. They have just bet a price 2.7% shorter than the one they believed they were taking. On a €200 stake, that is €5.40 gone. Not catastrophic on one bet. Catastrophic across a portfolio if it happens repeatedly, which it does, because live markets move every few seconds and the toggle reverts every three.

The 31% figure matters because it tells us this is not a niche edge case. It is the behaviour of a third of the users who engage with the feature. And the feature is not obscure. Format toggles are standard on every major Croatian-facing sportsbook — Sofascore's odds widget links out to them, tipster Telegram channels post in fractional, and a generation of Croatian bettors who grew up on British football coverage think in 5/4 and 11/10 rather than 2.25 and 2.10. The toggle is not a power-user tool. It is mainstream.

What the Croatian regulatory picture says — and doesn't

Croatia's online betting market is regulated under the Zakon o igrama na sreću, with Hrvatska lutrija d.o.o. acting as the licensing and oversight body for most online operators since the 2015 amendments opened the market to EU-licensed books. The framework requires operators to display odds in decimal format, to show the potential return before bet confirmation, and to maintain records of the odds at the time of bet acceptance. What it does not do — at least not explicitly — is regulate how a display toggle behaves when a user switches formats and switches back.

That is the gap. The regulation assumes the displayed odds are the odds. It does not anticipate a scenario where the display format itself is a moving part that can change the price on the slip without the user's input. An operator could argue, with some justification, that the decimal price shown at confirmation is the binding price and the user had every opportunity to see it. A regulator could counter that the user saw a different number two seconds earlier and was not warned. Both positions are defensible. Neither has been tested.

The practical consequence is that Croatian bettors using the toggle are in a grey zone. They are not being defrauded in any provable sense — the decimal price is always shown, always accurate, always the price they get. But they are being nudged into accepting a price they did not consciously choose, and the nudge is baked into the interface by design. That is a different kind of problem from a rigged game or a refused withdrawal. It is quieter, and it scales.

The responsible gambling angle nobody mentions

There is a responsible gambling dimension here that gets lost because it does not look like a responsible gambling issue. Problem gambling interventions typically focus on stake limits, session timers, and self-exclusion. All of those are important. But a user who is repeatedly accepting prices 2-3% shorter than they intended is losing money faster than their stake size suggests, and the loss is invisible to them. They think they are betting €50 at 1.85. They are betting €50 at 1.78. Over a session, that difference compounds, and the user has no feedback loop telling them something is wrong.

If you are using a format toggle on a Croatian book, the practical mitigation is simple: after any toggle revert, re-check the decimal price on the slip before confirming. If it has moved, decide whether you still want the bet at the new number. That is 10 seconds of attention that saves the 2-3% leak. It is not a fix for the interface problem, but it is a fix for your bankroll. And if you are betting more than you can comfortably lose because the numbers on screen feel familiar, that is worth pausing on regardless of what format they are in.

How to test whether your book does this

You do not need to take my word for the 31% figure or the 3-second default. You can check your own book in about two minutes.

Open a live market — in-play football or basketball, something with frequent line movement. Toggle to fractional or American. Watch the clock. If the format reverts on its own, note how long it takes. Three seconds is the most common default, but some books use five, and a few use fifteen. Then watch the decimal price on the slip. If it changes at the moment of reversion, or within a second or two after, the book is recalculating your slip against the current line rather than the line you selected. That is the behaviour that produces the 31% mispricing.

A book that handles this correctly will either keep the format until you change it back, or show a "price changed" prompt when the line moves. A book that handles it badly will silently update the slip and let you confirm at the new price. Most Croatian-facing books are in the second category, not because they are malicious but because nobody flagged the interaction between the toggle timer and the slip recalculation during QA. It is a product oversight, not a conspiracy. But the effect on your returns is the same either way.

The fix operators could ship tomorrow

The obvious fix is to decouple the display format from the slip calculation. The slip should always lock the decimal price at the moment of selection, regardless of what format is displayed. If the line moves, the slip should prompt, exactly as it does when you leave a bet unconfirmed for too long. That is a two-line change in most betting engines. The reason it has not shipped is that the toggle timer was never treated as a pricing event in the first place — it was filed under UI, and UI changes do not go through the same review as pricing logic.

A second fix is to remove the auto-revert entirely. Let the user keep the format they chose until they change it. The cognitive load argument does not hold up when the alternative is a silent repricing. If a user wants decimal back, they can tap the toggle. That is one tap. The current setup saves them that tap and costs them 2-3% on a third of their toggled bets. The trade is not close.

A third option, and the one I would push for if I were advising a Croatian operator, is to log every toggle revert that coincides with a slip price change and surface it in the bet history. If the user can see, after the fact, that the price they confirmed was different from the price they saw, they can adjust. Right now that data exists on the operator's side — the odds history is recorded for regulatory purposes — but it is never shown to the player. Making it visible would cost almost nothing and would resolve most of the trust problem.

The number that should worry operators more than regulators

The 31% figure is the headline, but the number that should worry operators is the retention curve. A bettor who discovers, six months in, that a chunk of their losses came from prices they did not consciously accept does not just stop using the toggle. They stop trusting the book. Croatian bettors are not short of options — there are more than a dozen licensed operators competing for the same audience, and switching costs are close to zero. The operator that ships a clean toggle, or no toggle at all, has a genuine differentiator that costs nothing to implement and is almost impossible for a competitor to copy without admitting the problem exists.

The counterargument is that most users never notice. That is true. Most users also never notice a 2% house edge difference between two slots, and yet the entire slot market competes on RTP. The difference is that RTP is disclosed. The toggle repricing is not. That asymmetry — where the operator knows the price changed and the user does not — is the part that becomes a regulatory issue if anyone bothers to test it properly. Hrvatska lutrija has not published guidance on toggle behaviour, and no Croatian operator has been fined over it. That could mean the issue is not on the radar. It could also mean nobody has filed a complaint specific enough to trigger a review.

If you are a Croatian bettor who uses a format toggle, the implication is straightforward: check the decimal price on your slip after every revert, and if your book silently updates it, treat that as information about how the book is built. If you are an operator, the implication is less comfortable. The 3-second timer is a default you probably inherited from a white-label provider and never examined. The 31% is a behaviour your users are exhibiting right now, on your platform, without either of you talking about it.

The open question is whether the first Croatian operator to publish toggle behaviour data — or the first regulator to ask for it — forces the rest of the market to follow. If that happens, the 3-second revert becomes a case study in how a UI default turned into a pricing policy nobody voted for. If it does not, the 31% keeps betting the old price, and the old price keeps not existing.