Spunk Odds Reference

Overview

House edge and expected value

One calculation sits underneath every page of this site: the chance of a result multiplied by what it pays. This page explains it once, with numbers for each format.

Fair odds and offered odds

If an event has probability p, a fair bet on it returns 1 / p times the stake when it happens. On average the player then gets back exactly what was staked. A game makes money by paying less than that. If the offered return is (1 - e) / p, the player gets back 1 - e of each stake on average, and e is the house edge.

Example: a dice bet with a 25% chance has a fair return of 4.00x. With a 1% edge the offered return is 3.96x. The expected net result per unit is 0.25 x 3.96 - 1 = -1.00%.

The same edge in different formats

One example bet per format, all built with a 1% edge in this site's models
Format and betWin chanceTotal return on a winEV per unit
Coin flip, one call50.00%1.98x-1.00%
Dice, 10% chance10.00%9.90x-1.00%
Crash, 3x target33.00%3.00x-1.00%
Hi-lo, face card 9, call higher30.77%3.22x-1.00%
Mines, 3 mines, 4 reveals57.83%1.71x-1.00%

Keno, plinko, wheel and scratch cards set their edge through a pay table instead of a single formula; their overview pages compute the return of example tables.

The edge is charged on every round

Expected loss grows in a straight line with the number of rounds, while the typical spread around it grows only with the square root. That is why short sessions are dominated by luck and long sessions by the edge. The table shows a coin-flip bet with a 1% edge.

Bet of 1 unit(s), win chance 50.00%, total return 1.9800x on a win
RoundsExpected netSD of netRough 95% band (normal approx.)
10-0.103.13-6.36 to 6.16
100-1.009.90-20.80 to 18.80
1,000-10.0031.31-72.61 to 52.61
10,000-100.0099.00-298.00 to 98.00

After 10 rounds the expected loss of 0.10 units is lost in a spread of about 3.13. After 10,000 rounds the expected loss is 100 units and the spread is about 99: the result is now almost certainly negative. The rough 95% band in the last row lies entirely below zero.

Turnover, not stake

A useful way to think about cost is turnover: the total amount bet. Wins are usually re-staked, so a small bankroll can produce a large turnover. Someone who starts with 100 units and keeps betting one unit per round on a 1% edge game will, on average, have lost 100 units by the time the turnover reaches 10,000 units, if the bankroll lasts that long. The edge applies to the turnover, not to the money brought to the session.

Why no betting system removes the edge

Each round has a negative expected value. A betting system only changes how much is staked on which round. A sum of negative numbers, each multiplied by a positive stake, is still negative. Systems such as doubling after a loss make frequent small wins more likely but add a rare, very large loss. With a 49.50% win chance, the chance of ten losses in a row is 0.108%, and covering those ten steps needs 1,023 units of bankroll to win one.

Return to player

Some services describe a game by its return to player (RTP) rather than its edge. They are two sides of one number: RTP = 1 - edge. A 99% RTP is a 1% edge; a 96% RTP is a 4% edge, four times as costly per unit of turnover. When a pay table is complex, the RTP is the expected value of the table plus one, and the format overview pages show how to compute it line by line.

Each game guide repeats this arithmetic for its own format and parameters. If you gamble, the most useful number to know before you start is the edge, and the second most useful is the limit you set for yourself (see the responsible gambling page).