Game events are translation of the randomly generated floats into a relatable outcome that is
game specific. This includes anything from the outcome of a dice roll to the order of the cards
in a deck, or even the location of every bomb in a game of mines.
Below is a detailed explanation as to how we translate floats into events for each particular
different game on our platform.
Blackjack, Hilo & Baccarat
In a standard deck of cards, there are 52 unique possible outcomes. When it comes to playing
Blackjack, Hilo & Baccarat on our platform, we utilise an unlimited amount of decks when
generating the game event, and therefore each turn of a card always has the same probability. To
calculate this, we multiply each randomly generated float by 52, and then translate that result
into a particular card, based on the following index:
The only differentiating factor involved with these games is that with Hilo and
Blackjack
there is a curser of 13 to generate 52 possible game events for cases where a large amount of cards
are required to be dealt to the player, whereas
when it comes to Baccarat we only ever need 6 game events generated to cover the
most amount of playable cards possible.
Diamond Poker
When playing Diamond Poker, there is 7 possible outcomes in the form of gems. To achieve this,
we multiply each float generated by 7 before it is translated into a corresponding gem using the
following index:
// Index of 0 to 6 : green to blue
const GEMS = [ green, purple, yellow, red, cyan, orange, blue ];
// Game event translation
const gem = GEMS[Math.floor(float * 7)];
Both the dealer and the player are dealt 5 gems each,
which means that a complete game of Diamond Poker requires the generation of 10 game events. The
first 5 are assigned to the dealer and the second 5 are assigned to the player.
Diamonds
When playing Diamonds, there is 7 possible outcomes in the form of gems. To achieve this, we
multiply each float generated by 7 before it is translated into a corresponding gem using the
following index:
// Index of 0 to 6 : green to blue
const GEMS = [ green, purple, yellow, red, cyan, orange, blue ];
// Game event translation
const gem = GEMS[Math.floor(float * 7)];
The player is then dealt 5 gems.
Dice Roll
In our version of dice, we cover a possible roll spread of 00.00 to 100.00, which has a range of
10,001 possible outcomes. The game event translation is done by multiplying the float by number
of possible outcomes and then dividing by 100 so that the resulting number fits the constraints
of our stated dice range.
// Game event translation
const roll = (float * 10001) / 100;
Limbo
When it comes to Limbo, we use a two-step process. Firstly, we take the float and multiply it by
both the maximum possible multiplier and the house edge. Then, in order to generate a game event
that has
probability distribution
, we divide the maximum possible multiplier by the result of the first step to create the game event
in the form of a crash point.
// Game event translation with houseEdge of 0.99 (1%)
const floatPoint = 1e8 / (float * 1e8) * houseEdge;
// Crash point rounded down to required denominator
const crashPoint = Math.floor(floatPoint * 100) / 100;
// Consolidate all crash points below 1
const result = Math.max(crashPoint, 1);
Plinko
For any game of Plinko, the generated outcome is based on the path of the falling ball. The game
event determines the direction of the falling ball for each level in the falling process.
Players can choose between 8 and 16 pins of play,
which determines the number of game events required to generate a complete path from top to
bottom. Since there are only two possible directions (left or right) the translation is done by
multiplying each float by 2, which maps to the following index:
// Index of 0 to 1 : left to right
const DIRECTIONS = [ left, right ];
// Game event translation
const direction = CARDS[Math.floor(float * 2)];
Roulette Roll
Our Roulette is derived from the European version of the game where the wheel consists of 37
possible different pockets, ranging from 0 to 36. The game event is calculated by multiplying
the float by 37 and then translated into a corresponding pocket using the following index:
Traditional Keno games require the selection of 10 possible game events in the form of hits on a
board. To achieve this, we multiply each float by the number of possible unique squares that
exist. Once a hit has been placed, it cannot be chosen again, which changes the pool size of the
possible outcomes. This is done by subtracting the size of possible maximum outcomes by 1 for
each iteration of game event result generated using the corresponding float provided, using the
following index:
The fisher-yates shuffle implementation is utilised to prevent duplicate possible hits being
generated.
Mines
A mine game is generated with 24 separate game events, in the form of mines on the board. Each
float is multiplied by the number of possible unique tiles still remaining on the board. This is
done by subtracting the number of tiles remaining by 1 for each iteration of game event result
generated using the corresponding float provided. The location of the mine is plotted using a
grid position from left to right, top to bottom.
The fisher-yates shuffle implementation is utilised to prevent duplicate possible hits being
generated. Between 1 and 24 game event results are used, based on the settings chosen.
Video Poker
A video poker game involves 52 separate game events, in the form of cards in a deck. Each float
is multiplied by the number of possible cards still remaining in the deck. This is done by
subtracting the number of cards remaining by 1 for each iteration of game event result generated
using the corresponding float provided. This is done by selecting a card from the following
index:
The fisher-yates shuffle implementation is utilised to prevent duplicate cards being generated.
Wheel
The game event number is calculated by multiplying the float by the possible outcomes in the
segment. It is then used to determine the game event result as a multiplier, using the following
index:
See the
BitcoinTalk seeding thread to learn about how we utilise the salt hash based provable fairness modal for
this particular game.
Slide
See the
BitcoinTalk seeding thread to learn about how we utilise the salt hash based provable fairness modal for
this particular game.
Scarab Spin / Tome of Life
The game event number is calculated by multiplying the float by the possible outcomes in the
reel. The first 4 reels have a length of 30 possible outcomes, whilst the last reel has 41. The
game event determines the central stop position for each reel. This game consists of 5 game
event numbers, until the case of a bonus round, where more are generated.
Blue Samurai
Blue Samurai slots has 3 different types of spins. Regular, bonus and special.
For regular and bonus spins, 18 floats from 0 to 1 are generated from your hash. Unlike Scarab
Spin slots, which has fixed reels, Samurai slots has dynamic reels, meaning each symbol is
generated from the corresponding float that was assigned to it.
We use weighted random sampling to assign each float to its corresponding tile, in the same
order, moving down the reels, from left to right. Each symbol has its own fixed probability /
chance of appearing in any one tile, with the outer 2 reels having a different set of
probabilities to the inner 3 reels. For a bit more information on how symbols are selected, see
fitness proportionate selection algorithm to learn more.
Special spins are slightly different. For a start only 12 floats are taken from your hash, as
the outer reels are disabled. Between each special spin, any samurai symbols stay in place for
the remainder of the game, with the result being the final
count of samurais. This means that if you were to have for example 1 samurai stick in the first spin
- we'd technically only need 11 floats for the subsequent spin. For the sake of simplicity in the
probably fair model, we just generate 12 floats every time, and if the float that was allocated for
a tile has a stuck samurai from a previous spin, then that float is not used at all.
Dragon Tower
A Dragon Tower game is generated with 9 separate game events, in the form of the levels up the
tower. We generate a number of eggs depending on the difficulty for each level, and have a range
of tiles the egg can be on also represented by an integer.
Each float generated is then converted to integers to determine the egg location on each row.
For example: A level on difficulty easy would be represented like this: [0, 1, 3] - eggs would
be present at tile 1 and 2 and 4.
// count represents the number of eggs
// size represents the number of possible squares
const LEVEL_MAP = {
easy: { count: 3, size: 4 },
medium: { count: 2, size: 3 },
hard: { count: 1, size: 2 },
expert: { count1, size: 3 },
master: { count: 1, size: 4 },
}
The fisher-yates shuffle implementation is utilised to prevent duplicate eggs on a row.
