Skip to content
68tests680,000spins open data closed rig 18+

Research · articles from the lab

What the tests say when you put them together

Every review on this site is one measured run. The articles here read across all of them: 68 tests, 680,000 logged rounds, 2,900 bonus purchases, one method. Where an article quotes the test set, its figures are not written once; they are recalculated from the current tests every time the page is opened.

  • official demo servers
  • live figures, dated
  • per-test logs downloadable
dataset behind the articles68 tests · one cell each
latest testWild Beach Party · 144.9% · 2026-09-16
  • 68slot tests
  • 680,000rounds logged
  • 2,900bonus purchases
  • 2,705natural bonuses
3 articles publisheddata as of 2026-09-16

Articles

newest first
  1. How Often Do Free Spins Trigger? Natural Bonus Frequency, Measured

    No game publishes how often its free spins land. This report counts: every natural trigger in ten thousand logged rounds per game, across the whole lab, with the intervals, the longest waits, and the reason an average hides the part that decides a session.

    Lab report published 2026-09-17 data as of 2026-09-16 68 tests 4 min read
  2. Is Buying the Bonus Worth It? Fifty Purchases per Game Across the Lab

    Bonus buy buttons promise the feature without the wait, at a price of a hundred bets or more. This report reads fifty purchases per game across every tested slot that offers one: what the average ticket returned, what the typical ticket returned, and how many tickets lost money.

    Lab report published 2026-09-17 data as of 2026-09-16 68 tests 4 min read
  3. Declared vs Measured RTP: What the Lab’s Tests Show

    Every slot prints an RTP on its help screen. This report puts that figure next to what ten thousand logged rounds per game actually returned, for every test in the lab, and explains why the two rarely match and what the gap can and cannot tell you.

    Lab report published 2026-09-17 data as of 2026-09-16 68 tests 5 min read

Coming next

a topic is written up once the sample is big enough to say something
  • Q01Dead streaks: what the volatility label actually deliveredlongest dry runs and drawdowns against the studio label
  • Q02Max win reality: how close any test came to the advertised caplargest rounds as a share of the cap, across the lab
  • Q031000x editions against the originalsGates, Sweet Bonanza, Sugar Rush, Starlight, Athena side by side
  • Q04The Big Bass family: ten fishermen, one collector mechaniccollection share, retriggers and purchase value per edition
  • Q05Megaways, clusters and lines: hit rate by grid typewhat the mechanic predicts and what it does not
  • Q06Re-tests: how much a second 10,000-spin run moves the numberthe first games tested twice, run against run

How these articles are made

same rules as the reviews

A research article takes one question and answers it from the whole test set instead of one game. Its tables and charts are generated from the published tests at the moment the page loads, so they include every review added since the article was written, and the date next to each block says which test set produced it. The prose is written once and describes the shape of the evidence: which way the numbers lean and why. When a new batch of tests changes that shape, the text is revised and the revision is dated. Other articles here explain the mechanics behind the numbers without a live dataset; they say so at the top. The procedure behind every test is on the how we test page.

  1. 01One question per article. RTP, purchase value, trigger rate, dry runs: each gets its own page.
  2. 02Numbers from the tests, never typed. Every figure resolves to the current test set and carries its date.
  3. 03Sample size stated. Ten thousand spins per game is a lot for a review and little for a law; the reports say which is which.
  4. 04Verifiable by anyone. Each review links the spin log its figures come from, with a checksum.