What casinos hide. Part 2: Volatility

Understanding volatility: why a million spins might still not be enough

Last time we explored the odds of hitting a max win. Now let’s look at something casinos are even more reluctant to discuss: slot volatility.

The most common answer you’ll hear is:

“If volatility is low, you win often but small amounts; if it’s high, you win rarely but big.”

A perfect example of this is a screenshot from one of the most volatile and popular games out there — DOA2 (Dead or Alive 2).

Supposedly to make it easier for us, providers use vague adjectives instead of actual numbers: high, low, extreme, medium, and so on.

Why is this a problem?

First, you can’t even compare “high” to “high.” Two slots from the same provider might both say “high volatility,” yet differ by three or four times if you actually look at the numbers.

Second, these labels are basically useless.

It is like if every bank just said “low interest rate,” without ever telling you if that means 2% or 20%.

Just fluffy words with no actual figures. That is exactly what they want.

I see this as pure manipulation. I want to know exactly how often and how much, and I want to be able to compare real numbers, not vague adjectives.

The math behind it: standard deviation as the real measure of volatility

Fortunately, in some places we are finally seeing real numbers.

For example, on NoLimit’s site they actually list the standard deviation (σ). This is not just pretty marketing words, it is a concrete way to show the spread.

What do these numbers mean?

So what are you supposed to do with these figures? How can you tell if 18 is a lot or a little, and why does that get labeled as “extreme volatility”?

It does not even sound that extreme. Let’s break it down.

In classical probability, a standard deviation (σ) of 18 means that your result for a single trial typically fluctuates by about ±18 units around the expected value.

If your bet is $1, that would roughly translate to outcomes ranging from –$17.04 to +$18.96.

Of course in practice you cannot lose $17 on a single $1 spin. This is because a slot’s payout distribution is not actually normal. That is why it makes much more sense to analyze results over large distances, like tens of thousands of spins, rather than just looking at individual spins.

The Central Limit Theorem (CLT) guarantees that the final distribution will become roughly normal only with a very large number of spins. Besides, nobody walks into a casino to make just one spin.