There are 2 factors that determine how a game pays back to a player; RTP and variance. In this post we will be discussing both a games built in variance and factors that you can control as a player. Working closely in conjunction with RTP (which defines the average amount a player will win back from their staking over time), variance determines the manner which it is paid back. Both payout amounts and frequency will be influenced by a games volatility.
Lets take some well known slots as an example:
Here we have 2 slots which if you just look at RTP are very similar however they are anything but. Starburst is a slot that will give you lots of small wins (quite often less than your stake) with the very occasional medium win. Dead or Alive 2 is the king of high variance slots and could quite easily give you 100 dead spins in a row but has the potential for a max win of over 100,000 times the stake. 1000 people playing Starburst might average -40%/+40% of their initial deposit when finished with no massive disparity between them. 999 people could average -90% of their initial deposit on Dead or Alive 2 and 1 person could come out with 10,000% their deposit. Both slots have returned roughly the same amount but the distribution is massively different.
Slots rely on a random number generator (RNG) which will create a random number from a huge range and then translates that number into the end result and payout we see.
Nobody except the programmers responsible for a particular slot will know exactly how this RNG correlates to a slot as the logic will be unique to a slot or group of slots created by the same company but a question often asked is 'if the number generated is random, how can a slot have a fixed RTP?'. The answer to this is that the number is random but the programming logic behind the slot is not, this combined with the fact that RTP is measured over many millions of spins means that each slot with its unique rules and way of working is guaranteed over a long enough period of time to hit or be close to its RTP figure.
Take this example purely as that because I have never been involved with the programming of a real slot but as a programmer if asked to code such a system for a medium variance slot given a possible number range from 0-10,000,000, I would probably approach it something like this:
This is a very basic example with rounded numbers for ease of demonstration, it is likely that a real slot would have a lot more conditions before determining payouts but hopefully it gives you a general idea of how such a system could work.
The variance of table games is easier to understand than slots because we can see the underlying cause. In roulette variance is caused by the randomness of where the ball lands, in blackjack it is caused by the random cards received by both dealer and player.
Besides going and actually playing these casino games there is one incredible useful exercise that I would like you to do using the EVCalc simulator. Setup a new simulation like so (that's a cashable bonus, £100 deposit, £100 bonus, £3,000 wagering on a 97% RTP high slot doing £2.50 spins) and then set the number of runs to just 1. What this is going to do is now instead of calculating the EV of this offer by getting the average of a large number of runs, the EV shown is going to be representative of just a single attempt.
Run the simulation a few times and see the huge range that you will get. There will be a lot of -£100 and then a number of wins, in my tests I had: £1240, £213, £245, £36, £409, £779, £662. This also serves as a good example of how losing runs on high risk offers can look.
First we should establish a very important fact, you cannot change the variance/volatility of a game. What we can do is increase our exposure to this variance by playing a game more and therefore evening out any peaks and troughs in our wagering.