Odds Format Switch Adds 1.9s Before 36% Confirm Bets

A Croatian-facing sportsbook that switched its default odds display from decimal to fractional recorded a median delay of 1.9 seconds between the moment a bettor opened a selection and the moment they confirmed it, and 36% of those bettors abandoned the slip before confirmation. The figures come from a six-week A/B test run across roughly 41,000 sessions on a single operator's Croatian-language interface, split evenly between a control group on decimal odds and a treatment group defaulted to fractional. The headline number is small. The downstream effect on conversion is not, and it maps onto a problem Croatian operators have been slow to treat as a product issue rather than a customer-education one.

Where the 1.9 Seconds Comes From

The test was straightforward. New and returning users arriving on the sportsbook homepage were assigned to one of two display defaults. The control saw decimal odds (2.50, 1.85, 4.20). The treatment saw fractional odds (6/4, 17/20, 16/5). Both groups could switch manually at any point, and the switch was placed in the same position in the bet slip footer for both. Crucially, the operator did not change the underlying pricing, the margin, or the market depth — only the notation.

What the operator measured was time-to-confirm: the interval between a user tapping a selection and tapping the confirm button on the bet slip. In the decimal group the median was 4.3 seconds. In the fractional group it was 6.2 seconds. That 1.9-second gap is the number in the title, and it held across the full six weeks rather than washing out as users "got used to" the format.

Two details matter for anyone reading this as a Croatia-specific finding rather than a generic one.

First, the gap was wider among users aged 45 and over — 2.6 seconds — and narrower among users under 30, at 1.1 seconds. That is consistent with the idea that younger Croatian bettors have more exposure to fractional odds through UK-facing content and social media, while older users have spent their betting lives in decimal, which has been the dominant format in Croatian retail and online sportsbooks since the market regulated in 2015.

Second, the gap was almost entirely absent among users who had manually switched formats at least once in the preceding 30 days. Those users — about 14% of the treatment group — confirmed in 4.5 seconds, statistically indistinguishable from the decimal control. The problem is not fractional odds as such. It is fractional odds as an imposed default on people who did not choose them.

The abandonment number is the one that should worry operators

The 1.9-second delay is a curiosity. The 36% is a business problem.

In the control group, 22% of opened bet slips were abandoned before confirmation. In the fractional group, that rose to 36% — a 14-percentage-point increase, or a 64% relative jump in abandonment. The operator's own analysts flagged that some of this is likely noise from the A/B framework itself, since any interface change can depress conversion in the short term. But the effect persisted into weeks five and six, well past the point where novelty effects usually decay.

Stake sizes tell a complementary story. Among users who did confirm, the fractional group staked a median of €4.80 versus €6.10 in the decimal group. That is a 21% reduction in median stake, and it suggests the friction is not just slowing people down but making them less confident in what they are about to commit.

Why Croatia Is a Particularly Bad Market for This Experiment

Croatia's regulated market has a specific history with odds formats, and it explains why the fractional default performed worse here than similar tests have in other jurisdictions.

The Croatian Lottery (Hrvatska lutrija) and the retail sportsbook network that dominated the market before liberalisation both used decimal odds almost exclusively. When the online market opened up under the 2015 Gambling Act and the subsequent licensing regime administered by the Ministry of Finance, operators inherited a customer base that had been trained on decimal notation for decades. Fractional odds were never the local convention in the way they are in Ireland or the UK, where the format is embedded in how people talk about prices.

That means a Croatian operator defaulting to fractional is not offering a familiar alternative. It is asking a customer to do mental arithmetic in a format they may have seen but never used transactionally. The cognitive load is real, and the 1.9 seconds is the measurable cost of it.

There is a second Croatian-specific factor: the euro adoption in January 2023. Croatian bettors spent 2022 and early 2023 recalibrating every price they saw from kuna to euro, and the market's UX teams spent that period simplifying rather than complicating. A fractional-odds default lands badly in that context, because it reintroduces a translation step at exactly the moment users have been trained to expect fewer of them.

What the operator actually did next

The operator in question reverted the default to decimal within three weeks of the test ending and kept fractional as an opt-in only. That is the obvious call, and most Croatian operators reading these numbers would make it. But the more interesting question is what the test reveals about the assumption that odds format is a cosmetic preference.

If format were purely cosmetic, switching it should have produced a small, uniform, short-lived effect. Instead it produced a persistent 14-point abandonment gap and a measurable stake reduction. That suggests format is not cosmetic at all — it is part of the decision architecture, and treating it as a display setting rather than a conversion lever is a category error.

The Decimal Default Is Not Automatically Right Either

It would be a mistake to read this as "decimal good, fractional bad." The test only shows that in Croatia, in 2024–2025, on one operator's interface, a fractional default underperformed a decimal one. Several conditions could flip that.

For a Croatian operator targeting diaspora users in the UK and Ireland — a non-trivial segment given the size of the Croatian community in those markets — a fractional default might perform better, because that is the format those users encounter daily. The same logic applies to operators whose acquisition strategy leans heavily on UK-facing affiliate content, where fractional prices are the norm in editorial and odds-comparison pages.

There is also a generational drift worth watching. The under-30 cohort in the test showed only a 1.1-second gap, and that cohort is the one that will dominate the market in a decade. If fractional exposure through social media and cross-border content keeps rising, the Croatian decimal default could eventually become the friction point rather than the solution. No operator has run a long enough panel study to know where that crossover sits, and anyone claiming to know is guessing.

What the data does support is a narrower claim: defaulting users into a format they did not choose imposes a measurable cost, and the cost is larger in markets with a strong incumbent convention. Croatia has a strong decimal convention. The cost showed up.

The switch itself is a friction point, not just the format

One finding from the test deserves separate attention because it is easy to miss. Users who manually switched formats took a median of 11.4 seconds to complete the switch and confirm their bet — nearly three times the decimal baseline. The switch control was a small text toggle in the bet slip footer, and users frequently tapped it, read the result, tapped it again, and only then confirmed.

That is a UX failure independent of which format is default. If an operator is going to offer both, the switch needs to be legible, persistent, and ideally surfaced before the user reaches the bet slip, not inside it. A toggle that users cycle through three times before committing is worse than no toggle at all, because it converts a display preference into a decision the user has to make mid-transaction.

The operator's post-test fix addressed this too: the format selector moved to the account settings page and to a persistent header control on the sportsbook, with the current format always visible next to the balance. Early data from the follow-up period shows switch-related friction dropped to a median of 3.8 seconds, though the sample is still small.

What This Means for Bonus Economics

There is a less obvious implication here for anyone running acquisition bonuses in Croatia, and it is worth spelling out because it connects the UX finding to the numbers that actually drive operator P&L.

A 14-point increase in bet-slip abandonment is, functionally, a 14-point reduction in the rate at which deposited funds get wagered. For an operator running a standard welcome offer — say a 100% match up to €200 with a 35x wagering requirement on the bonus — that abandonment directly extends the time it takes a typical user to clear the requirement, and in some cases pushes them past the offer's 30-day expiry. Users who do not clear wagering do not convert to retained customers at the same rate, and the operator has paid the acquisition cost without recovering it.

The arithmetic is not precise, because abandonment is not uniform across stake levels and the operator in question did not publish cohort-level wagering data. But the direction is unambiguous. A format default that depresses confirmation rates and median stakes is, in effect, a tax on every bonus the operator runs. It is the kind of thing that never shows up in an affiliate dashboard, because the affiliate gets paid on the deposit, not on the wager.

For Croatian affiliates and content sites, the practical takeaway is to check what format the operator they are promoting actually defaults to, and to flag it if it is fractional. Most Croatian-facing operators default to decimal, and the ones that do not are usually doing so because their parent brand is UK- or Ireland-facing and the localisation was incomplete. That is a fixable problem, and pointing it out is more useful to readers than another welcome-offer comparison.

The Open Question: Is Format a Setting or a Signal?

The test leaves one thing unresolved, and it is the thing that matters most for how Croatian operators should think about this going forward.

Is odds format a neutral display setting that users should be able to choose freely, or is it a signal about who the product is for? The data here suggests the second. A fractional default on a Croatian interface reads, to a meaningful share of users, as a product that was not built for them — and the 36% abandonment rate is what that reads as in behavioural terms. It is not that those users cannot parse 6/4. It is that being asked to parse it, at the moment of commitment, on a site they expected to speak their language, costs the operator a bet.

The counter-argument is that format preferences are real and personal, and that a well-built product should accommodate them without penalty. That is true, but accommodation and defaulting are different things. Every operator in Croatia can offer both formats. The question the test raises is whether defaulting to the less familiar one is ever justified by anything other than the operator's own internal convenience — a shared codebase with a UK brand, a template that was never localised, a product decision made in a different market and inherited without review.

If the answer is that it is never justified, then the fix is trivial and most Croatian operators have already made it. If the answer is that it is sometimes justified, then someone needs to run the test in the specific context where it is supposed to pay off — diaspora acquisition, UK-facing affiliate traffic, or a younger cohort — and publish the numbers. Until that happens, the 1.9 seconds and the 36% stand as the only Croatian data points on the question, and they both point the same way.