A method is worth stating only if somebody else can run it

Everything on this site comes out of documents that were open to anybody, and our method amounts to reading them in a fixed order and recording which one each figure came from and on what day. There is no proprietary "scoring model" here, no set of criteria "weighted" out of a hundred, no "expert panel", and nothing at all that a careful reader with a free afternoon could not repeat from the same sources.

The claim is a modest one, and it is also the only claim this material will support, because the three things a reader would most want measured are every one of them out of reach from outside. A jackpot pool sits inside a supplier's server; the probability of its top level is a parameter of a game nobody outside the studio has ever seen, and the amount on the "counter" (the figure a page shows for a pool as it grows) is a number served by a script that answers to the supplier. None of it is open to measurement from where we sit, so measurement is not what these pages offer.

What can be done from outside is bibliography (the dull business of saying where a number came from), and bibliography done properly (dated, ranked, and open to contradiction) is worth a great deal more than a verdict done badly.

The order the sources are read in

The order below is followed for every figure on this site, and the first rank that holds the thing is the rank that gets cited. Five ranks, and no exceptions.

1. The supplier's own pages, press releases and game terms: Games Global, NetEnt and Evolution, Playtech, Blueprint Gaming, Relax Gaming, Red Tiger, Pragmatic Play.
2. Operators whose licence obliges them to describe a pool correctly in their own market: Grosvenor, BetMGM UK, PlayOJO, Playzee, Pink Casino, 21.co.uk. These are the best source for a seed in local currency (a seed is a nominal threshold per currency and never a conversion of another one).
3. The trade press, which reprints supplier statements and is dependable on amounts and dates: iGaming Business, Gaming Intelligence, European Gaming, CasinoBeats, G3.
4. The jackpot trackers and databases, dependable on trends and on the day a level last fell, and silent on seeds and on distribution: casinolistings.com, slotcatalog.
5. Comparison and affiliate sites, used only where nothing above holds the thing at all, and named as such ("according to a comparison site") in the line that carries the number.

The fifth rank is the one that matters, because it is where most of this trade takes everything and where we go last (and reluctantly). A figure from rank five is marked in the text rather than laundered into a "fact", and the useful finding is very often that ranks one to four are all silent: that silence is itself publishable, and on this site it gets published.

What happens when two good sources disagree

Good sources disagree. They do it far more often than a tidy page would suggest, and the rule here is simple: we print both, we date both, and we say the question is open rather than quietly choosing the one that reads better.

The return to player of Hall of Gods is 95.70 per cent on NetEnt's own page and 95.30 on the Evolution page for the same game (with a comparison site offering 95.5 for good measure), and the release date is 22 November 2010 in the first place and 2013 in the second.

Rank one settles both, which here means NetEnt, and the disagreement still earns a line of its own, because a reader who meets 95.30 somewhere else deserves to know why this site says something different. The largest Absolootly Mad prize was paid by Napoleon (the operator announced it) and by Spin Casino (according to a well-known jackpot site): one of those is a press release and the other is a mistake, and they are not two opinions of equal weight. The word "verified", in this trade, very often means "copied from a site that copied it".

Elsewhere no rank wins at all, and then the page carries the mess. The Dollar Ball record circulates as €100,808 of 8 July 2015, as $113,296 of 22 April 2014 and as the same sum four months later, which is three claims in two currencies; the top Age of the Gods win is €984,763 in one database and $1,186,695 of 19 December 2020 in another, and those cannot be the same event under any arithmetic.

The size of a "network" is no better served: the Jackpot King catalogue is "over 60" games at one licensed operator and "almost 40" at another, and Dream Drop is "over 50" titles in a Relax Gaming statement of December 2025 and "over 30" on a licensed casino's own page.

In each of those cases the page says what each source says, with the date it was read, and then stops there. A guide that closes an open question by choosing in silence has told you something it does not know!

Why there is no counter here, and no probability either

A "live" counter is the one thing a jackpot site is expected to carry. It was the first thing we refused. The figure is true for minutes; it cannot be checked afterwards by anybody, this site included, and the services that publish it state neither where it comes from nor how often it refreshes.

One Games Global title (King Cashalot, from the first wave of online progressives) has been sitting above $2.2 million since 2017 without paying anybody, which a ticker renders as excitement and a date (2017, which is nine years ago) renders as a warning.

What we put in its place is three figures that keep: the "seed" (the sum a level restarts from after it falls), the mean of every prize that game has really paid, and the "drought" (the gap since the last win) measured against the game's own average gap: three figures, all of them datable, and all of them still checkable a year from now.

The odds are the harder refusal, because readers ask for them directly and in exactly those words (the search is for the "chances of winning a progressive jackpot", and it is made by the thousand every month). Not one supplier publishes the probability of a top level (we have read the published terms of all four of the big ones): NetEnt, Games Global, Playtech and Relax Gaming all publish the return to player and the "hit frequency" (how often the base game pays anything at all), which is a different quantity describing a different event.

Every percentage in circulation is either unattributed or arithmetically incompatible with the recorded sequence of wins, and two of the most-quoted fail that test by an order of magnitude.

In place of a probability we give the observed interval, which is a count of real events with real dates attached to them. It is less satisfying than a percentage, and it is the only answer the published record can support (the alternative is to invent a denominator, which is what the circulating percentages amount to).

Why a card shows a licence and never a bonus figure

The reason is half regulatory and half editorial, and the regulatory half settles it on its own. In Ontario the Standard 2.05 of the AGCO puts inducements, bonuses and credits outside every channel except the operator's own site and direct marketing to a consenting player (an affiliate's description of an offer falls squarely inside the prohibition), Standard 2.06 requires the material conditions of an offer at first presentation, and Standard 1.21 obliges a registered operator not to work with affiliates who also promote unregistered sites open to Ontario players.

Alberta codified the same approach in June 2026 (in the gaming standards and in the first schedule of its regulation), four weeks before the market opened. A site built around offers therefore has to write itself twice, once for the readers it may show a figure to and once for the readers it may not; in practice most of them quietly write the single version that suits them, and hope.

The editorial half is simpler and would hold even without a regulator, since on a site about jackpots the thing under comparison is the "pool" (the fund the network holds) and a welcome figure tells you nothing whatever about whether a house carries the network that holds the pool you came for.

Every card here carries the regulator whose register we were able to read and the "contribution" of the networks the catalogue runs, which is information that stays true for months rather than for weeks, and the figures behind it are set out on how a pot grows.

What a date means on these pages, and what gets corrected

Every page carries the day its text was last revised, and that date moves only when the text has really been revised: an automatic substitution of the year (the usual way of "refreshing" a site) is how a guide ends up declaring that a market opened in 2026 or that an operator was founded last spring.

We read the figures from registers and supplier pages again whenever a page is revised, and figures that came from a tracker carry their own date inside the table that holds them (the column is headed with the day the tracker was read, not with the day the page went up).

Corrections are the part of any method that shows whether the rest of it was real. We change a wrong figure on the page rather than quietly in the archive, a source that turns out to belong to rank five is marked as rank five, and a game renamed by its own maker gets the name its maker uses (Icy Wonders and Temple Tumble 2 appear here under the names NetEnt and Relax Gaming actually use, not the two "titles" half the trade has invented for them). Send a correction to contact with the source attached, and it will be dealt with sooner than you probably expect!

Questions