Is counting cards illegal? What the count measures
Legal almost everywhere, grounds for a private ban almost everywhere, and useless against a reshuffled virtual shoe.
The subjects sit apart from the rest of this reference. One sets out how the main families of casino games are grouped: the ones settled by a wheel or a pair of dice, the ones settled by a shuffled deck, and the ones settled by a random number generator. It then explains why that grouping decides how a game behaves over a long session rather than over a single hand.
A second looks at three slot releases and at what a licensed title actually buys, which is an audience that already recognises the characters, and a set of mechanics that still has to work for a player to whom the licence means nothing.
The third is the one still read most often. It works through what a card count measures at a blackjack table: a running count, the correction that turns it into a true count, and why that correction matters more than the count it corrects. It then works through the reason none of it carries over to a table where the deck is reshuffled after every hand. That is the useful half, and it is arithmetic rather than opinion. A count means something only while cards are being removed from a finite deck and not replaced.
All three date from 2018, and the arithmetic in them has not moved since. House edge, return to player and variance are properties of a game’s rules, not of the year the rules were written down.
Three 2018 slot releases: Robin of Sherwood, Fortunium, Jumanji
Three slot releases from July 2018, described as they were, with the claims that could not be checked removed.
Online casino games, and how the four families differ
The blog that started here in 2018, rewritten as a plain description of what the category actually contains.
Common questions
What is the difference between house edge and return to player?
They are the same number stated from opposite ends. Return to player is the share of total stakes a game pays back over a very large number of rounds; the house edge is what is left. A game quoted at 96% RTP carries a 4% edge. Neither figure describes a session — both are limits that only appear over a volume of play no individual reaches, which is why a player can sit well above or well below either for a long time without anything being wrong with the game.
Why does variance matter more than the edge over a short session?
Because the edge is a rate and variance is a spread. Two games with an identical 4% edge behave completely differently if one pays small amounts often and the other pays large amounts rarely. Over a few hundred rounds the spread dominates the rate, so what a player actually experiences is variance; the edge only becomes the visible force over a horizon most sessions never reach.
Does a card count work against a continuous shuffling machine?
No, and the reason is the point of the third article. A count is a record of which cards have already left a finite deck, so it carries information only while the deck is being depleted. A machine that returns every played card to the shoe before the next hand restores the starting composition each time, so there is nothing to accumulate. The count is not weakened there; it is measuring nothing at all.
What does a licence actually buy on a slot release?
Recognition, and nothing else. A player knows the characters before the first spin, and the royalty for that has to be earned back. The mechanics underneath are judged on their own, because the same player stops thinking about the licence within a few minutes. That is why the second article treats the theme and the mathematics as two separate questions about one release.