Same Slot, Two RTPs: 31% Never Check the Help Tab

Around 31% of players never open a slot's paytable or help screen before spinning, according to operator-side behavioural data that has circulated in iGaming analytics circles for several years. That single habit is the reason two people can play the same title, at the same stake, on the same evening, and end up with materially different long-run returns. The gap isn't luck. It's a dropdown menu most people never touch.

The mechanism is unglamorous: many providers ship the same game in multiple RTP configurations, and the version you get is decided by the operator, not by you. A title like Book of Dead has been documented in the wild at 94.25%, 96.21%, and 96.57% depending on which regulated market and which operator you're on. Pragmatic Play's catalogue routinely appears at 94.5%, 96.5%, and occasionally 97%+. Play'n GO, Push Gaming, Hacksaw, and Relax have all shipped multi-RTP builds. The paytable screen — the "i" icon, the question mark, the hamburger menu in the corner — is the only place the live number is disclosed in-game. Skip it, and you're guessing.

Why the same slot exists in more than one RTP version

The short answer is margin. The slightly longer answer is that margin, licensing, and market segmentation all interact, and Croatia sits in the middle of that interaction.

A slot's return to player is the complement of the house edge over an infinite sample. At 96.5% RTP, the operator keeps 3.5% of turnover in expectation. At 94.5%, it keeps 5.5%. On a game turning over, say, €2 million a month across a mid-sized Croatian-facing operator, that two-point difference is roughly €40,000 per month, or close to half a million euro a year on one title. Multiply by a portfolio of 3,000 games and the incentive to run the lower build is obvious.

Providers don't usually frame it that way. The standard line is "market flexibility": regulators in some jurisdictions cap or floor RTP, tax regimes differ, and operators want a game that still works economically after a 15% or 20% gross gaming revenue levy. That's not entirely spin. Malta, the UK, Romania, and various Latin American markets have different effective tax loads on slots, and a 96.5% build that's comfortable under a 5% tax can be marginal under a 20% one. Croatia's own framework — online slots taxed under the games-of-chance regime rather than the sports betting regime — puts local operators in the higher-cost bucket, which is one reason the lower builds show up on .hr-facing sites more often than people expect.

What the flexibility argument doesn't justify is the disclosure gap. The RTP is technically available. It's just buried in a screen that a third of players never open.

The three-tier build pattern

In practice, most multi-RTP games come in three flavours:

• The "default" build, usually the highest, used in marketing material and provider spec sheets.
• The "operator" build, typically 1.5 to 2.5 points lower, which is what most regulated operators actually deploy.
• The "low" build, occasionally 3 to 4 points below the headline number, used in markets with punitive tax or where the operator is running aggressive bonus economics.

When a provider's website says "RTP: 96.5%", that's the default. When your casino's help tab says 94.25%, that's the truth for your session. Both statements are accurate about different products.

The 31% figure and what it actually measures

The 31% number comes from session-level telemetry: the share of player sessions where the paytable or help screen is never opened before the first spin. It's not a survey response, which matters, because self-reported behaviour on this question is wildly unreliable — ask players directly and you'll get 70%+ claiming they always check, which contradicts every piece of clickstream data operators have.

The figure varies by segment. Among players who have been active for more than three years, the non-check rate drops to roughly 22–24%. Among players acquired in the last 90 days, it climbs above 40%. Mobile sessions show a higher non-check rate than desktop, partly because the info icon is smaller and partly because the mobile UX flow tends to push you straight from lobby to reels with fewer interstitial screens. In Croatia specifically, where a large share of online slot play happens on phones during commutes and evening hours, the mobile-skewed non-check rate is likely closer to 38–42% than to the headline 31%.

There's a second-order effect that's more interesting than the headline. Of the players who do open the help screen, a meaningful fraction open it to check the paytable symbols or the bonus mechanic, not the RTP line. The RTP is usually on a separate tab — "Game Rules", "Return to Player", or in some cases a paragraph of legal text at the bottom of a long scroll. So even the 69% who "check" aren't necessarily checking the thing that matters for their expected loss.

What a two-point RTP difference costs

Run the numbers on a realistic Croatian player profile: €1 average stake, 600 spins per week, 50 weeks a year. That's €30,000 in annual turnover.

RTP build	Expected annual loss	Difference vs. 96.5%
96.5%	€1,050	—
95.0%	€1,500	+€450
94.25%	€1,725	+€675
92.0%	€2,400	+€1,350

The 94.25% build costs that player €675 more per year than the 96.5% build on identical stakes and identical spin counts. Over five years, that's €3,375 — enough to matter, not enough to notice in any single session. This is precisely why the disclosure gap survives: the cost is invisible at the session level and only shows up in aggregate, where no individual player is looking.

Variance does the rest of the work. A 94.25% game and a 96.5% game have identical volatility profiles in most cases — same reel strips, same paytable structure, same hit frequency. In a 500-spin session, the two builds are statistically indistinguishable. You cannot feel a 2.25-point RTP difference. You can only read it.

How to find the actual RTP on a Croatian-facing site

The information is there. It's just not where most people look.

In-game help screen. This is the authoritative source for your session. Open the menu, find "Game Rules", "Paytable", or the "i" icon, and scroll to the RTP line. It should state something like "Return to player: 94.25%" or "The theoretical RTP of this game is 96.21%". If the game shows a range — "RTP: 94.25%–96.57%" — that's a red flag, because it means the build varies and the screen isn't telling you which one you're on. Reputable implementations show a single number.

Operator game page. Many Croatian operators list RTP alongside volatility and max win on the game's landing page. This is more reliable than provider marketing but less reliable than the in-game screen, because operators occasionally forget to update the page when they switch builds.

Provider spec sheets. Useful for knowing the possible range, useless for knowing what you're actually playing. If a provider lists three RTP options, assume you're on the middle or lower one unless the in-game screen says otherwise.

Jurisdictional defaults. Some regulators mandate a minimum. Malta doesn't set a hard floor for slots. The UK Gambling Commission has pressured operators toward higher builds but hasn't legislated a number. Croatia's own rules don't specify a minimum slot RTP either, which means the in-game disclosure is effectively the only consumer protection on this specific point.

Red flags worth noting

• A game page that lists RTP as a range rather than a figure.
• A help screen that describes the RTP in vague terms ("high return") without a number.
• A provider whose marketing says 96.5% but whose in-game screen says 94.x% — this is legal, but it tells you something about the operator's priorities.
• Bonus terms that restrict which games count toward wagering and those games happen to be the lower-RTP builds. This combination shows up more often than it should.

Does the 31% figure justify regulatory intervention?

This is where the Croatian conversation gets interesting, because the country is small enough that operator behaviour is visible and concentrated enough that a rule change would actually bite.

The argument for mandating prominent in-game RTP disclosure — not buried in a help tab, but shown on the loading screen or in the game frame — is straightforward. The information asymmetry is total: the operator knows the build, the player doesn't, and the only place the player can find out is a screen that 31% never open. Under Croatian consumer protection law more broadly, that asymmetry would be treated as a disclosure failure in most other product categories. A bank selling two versions of the same savings product at different interest rates, and disclosing the rate only in a submenu, would attract regulatory attention.

The counter-argument is that RTP is disclosed, technically, and that players who care can find it. Slot play is entertainment spending, not investment, and the expected loss is the price of the entertainment. Mandating prominent RTP display might create a false impression that slots are a return-generating product, which they aren't — every build is negative EV, and the difference between 94% and 96.5% is the difference between two losing propositions, not between losing and winning.

Both arguments have merit, and the practical outcome in most markets has been a middle path: pressure on operators to raise builds and improve disclosure, without a hard legal mandate. The UK went this way. Malta has been slower. Croatia, which tends to follow EU-adjacent regulatory norms with a lag, hasn't moved on this specifically.

What operators actually do when pushed

When regulators or consumer pressure force the issue, operators tend to do one of three things. They raise the build across the portfolio — expensive, and rarely done fully. They raise the build on the most visible titles and leave the long tail at lower RTP — common, and hard to detect because nobody checks the long tail. Or they keep the builds and improve disclosure, betting that a small percentage of players will change behaviour and the rest won't notice. The 31% figure suggests that bet is usually correct.

The uncomfortable implication is that the non-check rate is itself the product. An operator running a 94.25% build on a game that the provider markets at 96.5% is relying on a specific, measurable behaviour — the failure to open a menu — to capture the difference. That's not fraud. It's segmentation by attention, and it works because attention is scarce and the cost of inattention is invisible in any single session.

What changes if you start checking

For the individual player, the practical effect of checking the help tab is narrow but real. You can't turn a negative-EV game positive. You can, over a year of regular play, reduce your expected loss by several hundred euro by choosing 96.5% builds over 94.25% builds on the same titles. You can also use the RTP line as a rough proxy for how the operator treats its players more broadly — an operator running low builds across the board is usually also running tighter bonus terms, slower withdrawal processing, and more aggressive wagering requirements.

There's a portfolio effect too. If you play ten different slots regularly and half of them are low builds, the blended RTP of your play drops by roughly a point, which on €30,000 annual turnover is about €300 a year. Not life-changing. Not nothing.

The more interesting question is whether the 31% figure is stable, rising, or falling. Operator telemetry suggests it's been roughly flat for five years, despite widespread industry discussion of RTP transparency. That flatness is itself informative: the players who check were always going to check, and the players who don't were never going to start, regardless of how much the topic gets written about. Which raises the question of whether transparency campaigns aimed at players can move the number at all — or whether the only lever that works is a regulatory one that removes the choice from the operator's hands and puts a single RTP on every game, in every market, with no help tab required.