Dice gambling glossary

#

1-3-2-6 system

A positive progression for even-money bets. Stakes go 1, 3, 2 and 6 units on consecutive wins, and the bettor returns to 1 unit after any loss or after the fourth win. Do dice betting systems work?

A

Anti-martingale

Any progression that raises the stake after a win and drops back after a loss, the mirror image of martingale. Losses stay small and profits come in bursts when several wins arrive in a row. Paroli is a capped version. Paroli script

Auto-Tuner

The DiceSim tool that runs a script many times with different params.* values, either as a grid search or with a genetic algorithm, over up to three parameters. Results are ranked by a fitness score of profit divided by drawdown, times the log of rolls survived. How to write a dice bot

B

Bankroll

The money set aside for a betting session or a strategy. Strategy risk is usually measured against it, for example as the longest losing streak the bankroll can pay for. Dice bankroll management

Base bet

The starting stake of a strategy, and the amount most progressions return to after a win. In a DiceSim script it is the basebet variable, often set as a fraction of the balance. Dice bankroll management

basebet (Lua)

Read/write Lua variable holding your base bet. The engine pre-fills it with the initial nextbet; most scripts set it at the top level and use nextbet = basebet to reset after a win. Lua reference

bethigh (Lua)

Read/write Lua variable that picks the side of the bet. false bets under and wins when the roll is below chance; true bets over and wins when the roll is above 100 - chance. Lua reference

C

Client seed

A string the player chooses or accepts, mixed into every roll of a provably fair game. Because the casino commits to its server seed before seeing the client seed, it cannot pick a server seed that produces chosen results for that pair. How provably fair dice works

currentstreak (Lua)

Read-only Lua variable with the current run of results: positive for wins in a row, negative for losses in a row. After six straight losses it is -6. Lua reference

D

D'Alembert

A linear progression: add one unit to the stake after a loss and take one unit off after a win, never going below the base bet. Stakes grow far more slowly than with martingale. D'Alembert script

Delayed escalation

A low chance strategy that bets the minimum through a preroll window of losses and only then starts multiplying the stake. DiceSim's template uses 3.35% chance, a 50-loss preroll and a 1.04 multiplier. Delayed escalation script

Dice bot

A program that places dice bets automatically by following rules written by the player, usually as a script. DiceSim runs dice bot scripts in Lua against simulated provably fair rolls. How to write a dice bot

dobet()

The Lua function every DiceSim script must define. The engine calls it after each roll, with win, balance and the other session variables already updated, and then reads nextbet, chance and bethigh for the next roll. How to write a dice bot

Drawdown

The fall in equity from its highest point. Maximum drawdown is the largest such fall during a run; DiceSim measures it on balance plus vaulted funds and reports it as a negative number. Reading simulation results

Dynamic recovery

A progression that sizes each bet after a loss so one win recovers the current drawdown plus one base bet: stake = (drawdown + base bet) / (payout - 1). DiceSim's template also caps the stake at a fraction of the balance. Dynamic recovery script

E

Efficiency score

Net profit divided by maximum drawdown for a run. A score of 2 means the run made twice as much as its worst peak-to-trough drop. The DiceSim leaderboard can rank runs by it. Reading simulation results

Expected value

The average result per bet over a very large number of bets. In dice it equals minus the house edge times the stake, so at a 1% edge every unit wagered has an expected loss of 0.01 units, whatever the win chance. Expected value and variance

F

Fibonacci system

A negative progression that follows the sequence 1, 1, 2, 3, 5, 8 and so on. The stake moves one step forward after a loss and two steps back after a win. Do dice betting systems work?

Flat betting

Placing the same stake on every roll. It gives the slowest and steadiest path toward the expected loss, and it is the baseline other systems are compared against. Do dice betting systems work?

G

Gambler's fallacy

The belief that a result becomes more likely because it has not happened for a while. Each dice roll comes from a fresh hash, so a win is no more likely after ten losses than after ten wins. Do dice betting systems work?

H

Hit rate

The share of bets that win. Over many rolls it settles near the win chance; a strategy can raise its hit rate by betting at a higher chance, which lowers the payout to match.

HMAC

Hash-based message authentication code: a hash function such as SHA-256 or SHA-512 applied to a message together with a secret key. Provably fair dice uses the server seed as the key and the client seed and nonce as the message. How provably fair dice works

House edge

The share of each wager the casino keeps on average, built into the payout. With a 1% edge, a 49.5% bet pays 2.00x instead of the fair 2.0202x. House edge explained

L

Labouchere

A cancellation system. The bettor writes a list of numbers, stakes the first plus the last, crosses both off after a win and appends the lost amount after a loss. Clearing the list banks its original total. Labouchere script

Losing streak

Consecutive losing bets. The chance that a given roll starts a streak of N losses is (1 - win chance)^N, and the longest streak a bankroll can pay for is the main limit on any negative progression. Dice calculator

Low chance dice

Betting at small win chances, usually 1% to 5%, for payouts of about 20x to 99x at a 1% edge. Wins are rare and dry spells of hundreds of rolls are normal. Low chance dice strategy

Lua

A small scripting language used by many dice bots. DiceSim runs Lua 5.4 in a sandbox with DiceBot-style globals such as chance, nextbet and dobet(). Lua reference

M

Martingale

Double the stake after every loss and go back to the base bet after a win. At a 2.00x payout each completed cycle wins one base bet, while a long enough losing streak loses the whole cycle. Martingale script

Max bet

The largest stake a casino accepts on one bet. It limits how many times a progression can raise its stake, however large the bankroll is. Do dice betting systems work?

N

nextbet (Lua)

Read/write Lua variable with the stake of the next roll. It must be a finite number of 0 or more, and a nextbet larger than balance ends the run as bankrupt. Lua reference

Nonce

A counter of bets made with the current seed pair. It starts at 0 and goes up by one per bet, so every bet gets a different HMAC message and a different roll. How provably fair dice works

O

Oscar's grind

A progression that aims to win one unit per series. The stake stays the same after a loss, goes up one unit after a win, and is cut so that a win never takes the series beyond one unit of profit. Do dice betting systems work?

P

params (Lua)

Read-only Lua table of values passed in from outside the script, such as from the Auto-Tuner. Scripts read them with a default, for example params.multiplier or 2. How to write a dice bot

Paroli

A positive progression that lets winnings ride for a fixed number of wins, usually three, then banks the run and starts again. Any loss costs one base bet. Paroli script

Payout multiplier

What a winning bet returns per unit staked, stake included: (100 - house edge) / win chance. At 1% edge, 49.5% pays 2.00x and 1% pays 99x. Dice calculator

Preroll

A number of losses a strategy waits through at the minimum stake before it starts a progression. It keeps turnover low while waiting for a long low chance drought to develop. Low chance dice strategy

previousbet (Lua)

Read-only Lua variable with the stake of the roll that just finished. nextbet = previousbet * 2 is the martingale step. Lua reference

Provably fair

A scheme in which each roll is computed from a server seed the casino committed to in advance, a client seed and a nonce. After the server seed is revealed, anyone can recompute every roll and check it. How provably fair dice works

R

Rakeback

A rebate some casinos and affiliate programs pay back to players, usually worked out from the amount wagered or from the house edge on it. It lowers the effective edge; the terms vary from site to site. House edge explained

Risk of ruin

The probability that a strategy loses its bankroll, or can no longer place its next bet, within a given number of rolls. DiceSim estimates it from the longest losing streak the bankroll survives. Risk of ruin explained

Roll

The result of one dice bet. In DiceSim and many crypto casinos it is a number from 0.00 to 100.00 with two decimals, derived from the first four bytes of an HMAC. How provably fair dice works

Roll over / roll under

The two sides of a dice bet. Roll under wins when the result is below the target, roll over when it is above it. In DiceSim both sides have the same win chance for a given chance. Lua reference

S

Seed rotation

Replacing the active server seed with a new one. The old seed is revealed so past bets can be verified, and the nonce restarts at 0. A DiceSim script can rotate with resetseed(). How provably fair dice works

Server seed

The secret random value the casino uses as the HMAC key for every roll. Its hash is published before betting and the seed itself is revealed after rotation, so players can check it was not changed. How provably fair dice works

Server seed hash

The hash of the server seed shown before any bet is placed, which commits the casino to that seed. DiceSim uses SHA-256 for it. After the reveal, hashing the seed must give the same value. How provably fair dice works

Standard deviation

The usual size of the swing around the expected result. For a single unit bet at a 1% edge it is close to 1 unit at 49.5% chance and close to 10 units at 1% chance. Expected value and variance

Stop loss

A rule that ends a session once losses reach a set amount. In a DiceSim script it is a check on profit followed by stop(). How to write a dice bot

Stress test

A DiceSim feature that runs the same script and settings on five workers at once, each with its own random seed pair. Only the dice change between runs, which helps separate luck from the strategy. Reading simulation results

T

Take profit

A rule that ends a session, or moves money to the vault, once profit reaches a target. In a DiceSim script it is a check on profit followed by vault() or stop(). How to write a dice bot

U

Unit

A fixed bet size used to describe stakes and bankrolls independently of currency, usually equal to the base bet. A bankroll of 1,023 units covers ten martingale losses from a 1-unit start. Dice bankroll management

V

Variance

How widely results spread around the expected value. Lower win chances and larger stakes raise it, so the same edge can produce very different short-term swings. Expected value and variance

Vault

Funds set aside so they cannot be bet. In a DiceSim script vault(amount) moves money out of balance; vaulted funds still count toward profit. How to write a dice bot

Verified run

A logged DiceSim run that the server replayed with the same script, seeds and parameters and found to match. Only runs the server can replay, up to 250,000 rolls, can hold a leaderboard rank. Leaderboard

W

Wagered

The total of all stakes placed, also called turnover. The expected loss of any strategy is the house edge times the amount wagered, which is why fast progressions lose more on average. House edge explained

Win chance

The probability of winning a bet, in percent. DiceSim accepts 0.01 to 98; a script sets it with the chance variable, and the payout multiplier follows from it and the house edge. Dice calculator

Where to go next