Slots tested on official demo servers. Every spin logged.

Q: What is Slotester Lab?

A slot-testing project that measures games instead of describing them. Software plays each game on its provider's official demo server for full recorded sessions, buys the bonus again and again where it is sold, and a human editor publishes the numbers with the log that produced them.

Q: Are demo servers the same as real-money play?

The demo build runs the provider's game math with virtual credits and reports the RTP profile of that build. What a specific casino adds on top, from bonuses to limits, is outside the test, and we cannot see what an operator serves to its own players.

Q: Why does the measured RTP differ from the declared one?

Because one run is one sample. Volatile games leave a wide confidence interval, and a single run can land above or below the declared figure without anything being wrong. Every review prints that interval next to the number.

Q: Can I check a review myself?

Yes. Each review links to its rounds file and prints the file's SHA-256 checksum. Two commands recompute the checksum and the RTP; the how-we-test page shows them.

Q: Do game studios or casinos pay for reviews?

Studios do not pay and cannot commission tests. Casinos can pay a commission if a reader follows a partner link under a demo; that link cannot change a measured number, a verdict or the order of reviews. The affiliate disclosure page has the details.

Q: Which games can be tested?

Any slot whose server protocol the rig speaks and whose official demo is public. Every game goes through the same fixed procedure, new reviews appear as the rig finishes them, newest first, and published games are retested with new runs over time.

Full recorded sessions and a series of bonus buys per game, retested over time, measured RTP with its confidence interval, dead streaks, drawdown and bankroll survival, next to the demo you can play yourself. The rig plays, a human editor checks, the spin log is yours to download.

See the tests How we test

latest runs spin 0

24 latest tests of 163, band = middle 80% · y = net result in bets, log scale

0slots tested
0rounds logged
0bonus buys
0natural bonuses

What every test measures

same procedure for every game, so the numbers compare

specimen Razor Shark Push Gaming · 10,000 rounds 01 / 12

Measured RTP

% of bet

82.8% 96.7%

Total win over total bet across every logged round of the test, runs pooled, with the 95% band that the sample size honestly allows. The white mark is the RTP the studio declares. The best RTP slots ranking sorts every test by this number.

Σ win ÷ Σ bet ± 1.96·sd/√n

Hit rate

% of rounds

24%

How often a round pays anything at all. The ten cells are ten rounds; the lit ones paid.

rounds with win > 0 ÷ rounds

Max win

x bet

1,016x

The biggest round of the test that did not trigger a bonus, on a log scale from 10x to 5,000x; the review adds the best round with a bonus, and the highest max win list ranks games by that bigger figure.

max(win ÷ bet), rounds without a bonus

Bonus buy

x price

—not sold in this build

A series of purchases at the listed price: the mean return against the price line, the median and how many returned less than they cost. Ranked on the best bonus slots page.

mean(return ÷ price)

Bonus frequency

bHz

4.2 bHz

Natural bonuses per 1,000 rounds, the lab’s own unit: the trace spans 1,000 rounds and every crest is one bonus. The review adds how long each one ran and what it paid.

natural bonuses ÷ rounds × 1,000

Dry run

rounds

31 rounds

The longest stretch of rounds without a single win, drawn between the two wins that framed it.

max consecutive win = 0

Drawdown and survival

bets

2,467 bets, deepest fall

The balance over the run in bets, the deepest fall from its running peak, and whether 50, 100 or 200 bets would have lasted the run.

max(peak − balance) ÷ bet

Every formula, the pipeline and the checks: how we test →

All tests together

163 tests · 282 runs · 2,820,000 rounds · updated with every new test

Measured against declared RTP

tests by gap, RTP points

153 / 163 within their own margin of error

beyond own margin: 10 · chance alone ≈ 8 below declared: 90 · at or above: 73

A single run of 10,000 rounds pins the RTP down to about ±17 points, so a gap of a few points is chance, not a verdict. More tests landed below the declared figure than above it: rare big wins lift the average, so the typical test sits slightly under it.

Mean measured RTP

95.9%

declared average: 96.38%
lowest test: 70.5%
highest test: 120.8%

One test is one sample: the hatched band is how far single tests spread around the declared figure, the white mark.

Bonus buys against the price

x price

40 / 113 above price

Each bar is one test: the average return of its bonus buys against the price, the white line. Across tests: 0.93x of the price, from 10,200 bonus buys.

How long a bankroll lasted

spins, log scale

50 bets 116 spins

100 bets 297 spins

200 bets 807 spins

Median spin at which a bankroll of that many bets ran out, all tests together. Only 1, 3 and 9 of 163 tests never exhausted 50, 100 and 200 bets.

Rhythm of the games

9–43% hit rate 114 rounds, longest dry run

From Tombstone R.I.P, where one round in 11 pays, to Sweet Bonanza, where 4 rounds in ten do. The longest dry run: Eye of Horus Megaways.

What a round pays, all tests together

share of 2,820,000 rounds by win in bet multiples · linear scale

88% of all rounds returned less than the bet. The rest is where every session is decided.

74% nothing
14% 0–1x
4.9% 1–2x
4.3% 2–5x
1.4% 5–10x
0.7% 10–20x
0.4% 20–50x
0.2% 50–100x
0.1% 100–500x
<0.1% 500x+

From the lab

all reports →

Lab report · 2026-10-03 11 min

Are Online Slots Rigged? The Same Slot Tested Twice, Run Against Run

The lab runs some slots more than once: same game, same demo server, same stake, a fresh session weeks later. If a game were tuned to pay less after a generous session, or to drift away from its declared RTP, the second run is where it would show. This article compares every re-tested slot run against run, checks whether the gaps are bigger than chance allows, and looks at what happens in the hundred spins after a big win.

Lab report · 2026-10-02 13 min

Max Win Reality: How Close Tested Slots Get to Their Advertised Cap

Every slot now ships with a max win printed above the reels, and the numbers keep growing: 10,000x, 50,000x, 150,000x. This article puts those ceilings next to the largest round that actually happened in each logged test, asks how rare a max win is by the studios' own odds, and finds out whether a bigger cap gives you bigger wins in a session or just a bigger number on the splash screen.

Lab report · 2026-09-29 10 min

Slot Volatility, Measured: What the Label Tells You and What It Does Not

Every slot carries a volatility label, and every guide to high or low volatility slots repeats it. This article checks the label against the logs: how often each labelled group paid, how long its dry runs lasted, how far a bankroll fell, and how wide a single round's swing really was. The label turns out to describe one thing well and two things not at all.

How the lab works

the rig plays · the editor checks · the log is public

01 Open the official demo The rig talks to the game server the way the game client does and records every response. No simulation, no emulated RNG. 11 studios · official demo servers
02 Play the standard run A full session of rounds at one bet, then a series of bonus purchases where the game sells them, then a real browser session for screenshots and video. Games are retested over time and every run stays on the page; the biggest round of each test feeds the slots with the highest maximum winnings list. 2,820,000 rounds · 10,200 buys · 282 runs
03 Compute, then check Every number comes from the log. Each round is reconciled against the server balance; one mismatch and the run is discarded. win = balance change, every round
04 A person signs off An editor reads the draft against the log and publishes it with the logs and their SHA-256 checksums. 163 reviews published

verify · Razor Shark

~/slotester $ sha256sum razor-shark-r1-rounds.csv 58ccfb2dade5e3425c8a5e7c5d2790c6ebaafbc33bbedae681acb498ede534ad razor-shark-r1-rounds.csv # ✓ the same checksum is printed on the review ~/slotester $ python -c "import csv;from decimal import Decimal as D;r=list(csv.DictReader(open('razor-shark-r1-rounds.csv',encoding='utf-8-sig')));print(round(100*sum(D(x['win']) for x in r)/sum(D(x['bet']) for x in r),2))" 82.76 # ✓ measured RTP of run 1 on the review: 82.76% ~/slotester $

Open data

Every review ships its logs with SHA-256 checksums.

Download the file, compare the checksum with the one on the page, recompute the RTP in one line. If your number differs from ours, write to the lab and the raw session is re-checked.

This log Its review How we test

different number? lab@slotester.com · about the lab

Questions

6 questions about the lab

What is Slotester Lab?

Lab answer

A slot-testing project that measures games instead of describing them. Software plays each game on its provider's official demo server for full recorded sessions, buys the bonus again and again where it is sold, and a human editor publishes the numbers with the log that produced them.

Are demo servers the same as real-money play?

Lab answer

The demo build runs the provider's game math with virtual credits and reports the RTP profile of that build. What a specific casino adds on top, from bonuses to limits, is outside the test, and we cannot see what an operator serves to its own players.

Why does the measured RTP differ from the declared one?

Lab answer

Because one run is one sample. Volatile games leave a wide confidence interval, and a single run can land above or below the declared figure without anything being wrong. Every review prints that interval next to the number.

Can I check a review myself?

Lab answer

Yes. Each review links to its rounds file and prints the file's SHA-256 checksum. Two commands recompute the checksum and the RTP; the how-we-test page shows them.

Do game studios or casinos pay for reviews?

Lab answer

Studios do not pay and cannot commission tests. Casinos can pay a commission if a reader follows a partner link under a demo; that link cannot change a measured number, a verdict or the order of reviews. The affiliate disclosure page has the details.

Which games can be tested?

Lab answer

Any slot whose server protocol the rig speaks and whose official demo is public. Every game goes through the same fixed procedure, new reviews appear as the rig finishes them, newest first, and published games are retested with new runs over time.

Slots tested on official demo servers. Every spin logged.

Slot catalog

Are Online Slots Rigged? The Same Slot Tested Twice, Run Against Run

Max Win Reality: How Close Tested Slots Get to Their Advertised Cap

Slot Volatility, Measured: What the Label Tells You and What It Does Not

Every review ships its logs with SHA-256 checksums.