How to Check Slot RTP: Find the Percentage in Your Game
You open a slot after reading that it has a 96% RTP. Its help screen says 94%. Which number belongs in your notes? The one tied to the version you actually opened, provided you have identified the mode and understood what the help screen is describing. A review can describe a different configuration without either number being a typing error.
The short answer: open the game's own information or rules menu, find the return-to-player entry, record the percentage together with the game, provider, mode and date, then investigate anything the screen leaves ambiguous. You do not need to place a bet to read a percentage. This guide takes you from locating that entry to writing a check that another reader could repeat.
What slot RTP measures — and what it leaves out
RTP means return to player: a theoretical relationship between amounts wagered and amounts returned under specified game rules. It is not the probability that your next spin wins. It is not the percentage of your deposit that will remain when you close the game. Nor does it tell you whether a particular feature will return its purchase price.
Three quantities are easy to mix up. Deposit is money transferred into an account. Turnover is the total amount wagered, including money that has already been returned and wagered again. Return is the amount paid by completed rounds. A €50 deposit can support more than €50 of turnover if some returns are played again; it can also disappear before reaching any particular turnover target.
The UK Gambling Commission's explanation of RTP distinguishes the long-run measure from individual play. Treat that as a definition, not as a promise about how long you should play. A losing sequence does not create a repayment schedule for your account.
RTP also leaves the distribution of outcomes unspecified. Two games can share a percentage while delivering returns in very different patterns. One might have frequent smaller payments; another might concentrate more of its theoretical return in less frequent features. A headline percentage alone cannot reconstruct that distribution.
A useful note also records where the figure was found: the game’s help panel, a studio factsheet or a third-party review. If two values differ, keep both with those labels while you investigate. Do not average them. A configuration difference is resolved by identifying which value belongs to the opened version, not by choosing a compromise percentage. If the help panel cannot be reached, the configuration remains unconfirmed even when the title and artwork match.
Find the game's help on a computer
Start inside the game window. The operator's navigation, account menu and promotions panel are different interfaces, even when they sit beside the reels. Look for an information symbol, a question mark, a menu button or a control labelled Help, Rules or Paytable. Labels vary, so use the control's wording rather than expecting every provider to place the same icon in the same corner.
Note the full game title and provider before opening the menu.
Open the information or rules control without starting a round.
Move through the help pages, including any final information page.
Look for RTP, theoretical return, return to player or payout percentage.
Read the sentence around the number, not just the number.
Record any version identifier and any distinction between base play and other modes.
On a phone: open the menu and check every rules page, including any scrollable section.
On a computer: inspect the information control near the game, not the operator’s promotional panel.
The Golden China rulebook, page 6, documents an information control. It is an example of a route into help, not evidence of the configuration used by a particular operator. Other titles may put their rules in a different menu. Read the complete help section before treating any percentage as the active RTP.
If the help opens in a separate panel, keep the game name in view or record it first. A tightly cropped percentage with no title or mode is surprisingly easy to mislabel later. Capture enough surrounding context to establish what you checked, while leaving account identifiers and balances out of any image you share publicly.
Find the same information on a phone
A mobile layout may collapse several controls into one menu. Rotate the device only if the game supports it and you can still read the labels. Do not assume that a desktop screenshot describes the exact position of a mobile button.
Open the game's menu and inspect the available entries before tapping. A settings gear may expose sound and speed controls while the rules sit elsewhere. If a panel scrolls, continue to its end: a short first screen is not proof that the information is missing. Where the browser offers text search, it can help with a text-based rule page, but it will not reliably search words drawn inside a game canvas.
Take two notes if necessary: the menu route and the information found. For example, “game menu → information → final rules page” is more reproducible than “bottom left.” Interface layouts change; a named sequence usually remains understandable.
If the text is too small, use supported zoom or open the linked official rules. A separate document is useful background, but label it as such. It does not automatically identify the configuration delivered by the game you launched.
One title can have several RTP configurations
A review often needs one number for a summary box. A provider's documentation can list several. That difference is exactly why checking the active version matters.
Golden China official rules, page 2: 88.01%, 94.01% and 96.01%.
These are alternative documented values, not three returns that a player can combine.
The source documents several RTP configurations. It does not tell you which one an unspecified operator has selected. Choosing the largest value would turn a genuine source into an unsupported claim. Match the version in your game’s information panel, or state that the active configuration remains unidentified.
The same Golden China symbol appears twice because visual identity and mathematical settings are different questions. Familiar artwork alone cannot identify the RTP. Record the actual entry available in your game and retain uncertainty if the help lists alternatives without naming the active one. The composition is not a pair of captured information panels.
This matters when comparing a demo with another version. A familiar logo, identical symbol artwork and matching feature names are useful identity clues, but they are not a substitute for checking the settings. The guide to what demo mode can establish walks through that comparison without treating a short demo run as a mathematical test.
Check whether the number belongs to your selected mode
Base play, a purchased feature and an optional enhanced mode are not interchangeable labels. Read the help for the option you are considering. A game may publish separate figures for different modes, one overall figure, or information that needs clarification. Do not fill a missing field by borrowing a number from another mode.
Suppose the rules list one RTP beside normal play and another beside a feature purchase. Your record should contain the label as well as the percentage. “96%” on its own loses the information needed to interpret it. “96% shown for normal play; purchased feature not checked” is narrower and more useful.
Also record the cost separately. RTP describes a ratio under rules; it does not tell you the amount debited when a feature begins. An option can have a different theoretical percentage and a much larger upfront price. The Bonus Buy cost guide explains why the base stake, purchase price and eventual return need three separate entries.
Do not activate or buy a mode merely to obtain a screenshot. Inspect its rules and confirmation information where available, then cancel if you are only researching. If the relevant information is unavailable without a transaction, mark that part unverified.
Compare 94% and 96% using turnover
Use a hypothetical €1,000 of turnover to keep the arithmetic visible. At 94%, the theoretical return associated with that turnover is €940; the corresponding theoretical loss is €60. At 96%, those figures are €960 and €40. The difference is €20 for the same €1,000 wagered.
The RTP difference is two percentage points. The theoretical loss falls from €60 to €40, which is one third lower. Those are different comparisons with different denominators. Neither says that a particular player will lose exactly €40 or €60.
Now change the turnover while keeping the percentages. At €100 wagered, the corresponding theoretical returns are €94 and €96. At €5,000 wagered, they are €4,700 and €4,800. Increasing turnover increases the amount exposed to the game's theoretical loss; it does not make continuing a losing session a recovery method.
A common mistake is to put the deposit into this calculation. If you deposit €100 and later wager a total of €350 through repeated play, €350 is the turnover figure. Your actual balance still depends on the actual rounds, not on multiplying either amount by the published RTP.
For a completed session, calculate observed return as actual payments divided by actual stakes. Keep that result in a different field. The session journal guide shows how to reconcile those amounts with the balance without renaming a short observation “the game's real RTP.”
Resolve a mismatch with a review or another screen
When two figures disagree, slow down before deciding that one is wrong. Compare the full title first. A sequel, a special edition and an original game can have very similar names. Then compare provider, mode, version information and the date of the source.
Work through the discrepancy in this order:
Does the review discuss the same title and edition?
Does it describe the same mode or just the default game?
Does it state a particular configuration, a range, or the highest listed RTP?
Does your current help screen identify an active value or merely list alternatives?
Is there a dated operator or provider clarification that resolves the difference?
If the review says 96% and the active help clearly identifies 94%, record both observations and their sources. Do not silently overwrite the live figure with the review. If the help lists 94% and 96% without specifying which applies, the correct entry is “active configuration not established.”
The same discipline applies to promotional screenshots. A release announcement can describe a feature without documenting every configuration offered through every operator. Our slot news section separates product announcements from the checks needed in an individual game instance.
Ask a question support can actually answer
A broad message such as “Is this game fair?” is unlikely to resolve a version mismatch. Ask for a specific fact and include the non-sensitive context needed to locate it. You can copy this outline and replace the bracketed fields:
I opened [full title] by [provider] on [date and time, with time zone]. In [mode], the help page shows [percentage or list of percentages]. A separate source shows [other figure, with URL]. Please confirm the RTP configuration of the game instance available through your site, the mode to which it applies, and where I can see that information in the game rules.
Add the device and the menu route if the issue is finding the information. If a support agent asks for a round identifier, use the operator's private support channel. Do not post account numbers, authentication details or unredacted personal screenshots in a public comment.
A reply that repeats a generic marketing percentage without identifying the configuration may leave the question unresolved. Keep the wording of the response and the date. You can accurately report “support stated X” without presenting it as an independent examination of the game software.
Keep a completed check, including what remains unknown
Here is a hypothetical record, not a report of testing an operator:
Game and provider: Example Slot, Example Studio.
Checked: 15 October 2026, 14:00, UTC.
Device and route: desktop browser; game menu → rules → information.
Mode: normal play.
Help entry: 94.00%, identified as the active configuration.
Review comparison: 96.00%, configuration not stated.
Other modes: not checked.
Evidence saved: game identity and relevant help entry, without account details.
Conclusion: use 94.00% for this recorded configuration; do not extend it to other versions.
The useful part is the boundary around the conclusion. You have established what the help states for a named configuration at a recorded time. You have not measured the mathematical model, predicted the next round or proved that another operator supplies the same version.
Can a few hundred spins verify the percentage?
No. A short observed return can sit far above or below a theoretical RTP. It cannot, by itself, identify which of several nearby configurations is running. Record the observation if you need it for accounting, but do not use it as a replacement for version information.
Does a higher RTP tell me when to play?
No. A comparison between theoretical percentages does not supply a timetable for favourable rounds. It also does not cancel differences in cost, availability or outcome distribution. Use the figure to understand the documented configuration, then make any decision about gambling within limits set independently of recent results.
What should I do if no active percentage can be established?
Write “not established,” keep the official material you found and request clarification. Leaving a field unresolved is more accurate than filling it with the most attractive number from a search result. For the next practical task — understanding how symbols become payments — continue to reading a slot paytable, or return to the slot guides index.
Discussion
Questions or something to add? Join the conversation.
Log in to join this discussion.
